Remade-With-Rust / rusty_h264

No fue posible recopilar el conjunto de dependencias resuelto para este informe: GitHub dependency-graph SBOM unavailable (404); the dependency graph may be disabled for this repository

{
  "data": {
    "repo": {
      "topics": [
        "built-with-rust",
        "codec",
        "ffmpeg",
        "rust",
        "embedded"
      ],
      "is_fork": false,
      "size_kb": 2081,
      "has_wiki": false,
      "homepage": "https://www.mata.network/remade-with-rust",
      "languages": {
        "Rust": 1565244,
        "Shell": 18421,
        "Python": 17109
      },
      "pushed_at": "2026-07-30T04:58:08Z",
      "created_at": "2026-06-24T18:11:38Z",
      "owner_type": "Organization",
      "updated_at": "2026-07-30T04:58:25Z",
      "description": "rusty_h264 is x264 remade with rust, focusing on embedded memory safe pixel encoding and decoding.",
      "is_archived": false,
      "is_disabled": false,
      "license_spdx": "BSD-2-Clause",
      "default_branch": "main",
      "license_spdx_raw": "BSD-2-Clause",
      "primary_language": "Rust",
      "significant_languages": [
        "Rust"
      ]
    },
    "owner": {
      "blog": "mata.network",
      "name": "Remade With Rust",
      "type": "Organization",
      "login": "Remade-With-Rust",
      "company": null,
      "location": "United States of America",
      "followers": 1,
      "avatar_url": "https://avatars.githubusercontent.com/u/295393398?v=4",
      "created_at": "2026-06-20T15:02:38Z",
      "is_verified": null,
      "public_repos": 11,
      "account_age_days": 39
    },
    "license": {
      "state": "standard",
      "spdx_id": "BSD-2-Clause",
      "raw_spdx": "BSD-2-Clause",
      "file_present": true,
      "scorecard_found": true,
      "profile_has_license": true
    },
    "activity": {
      "releases": [],
      "recent_commits": [
        {
          "oid": "b512bcb88f40dfad298251e274643fb95bcb359a",
          "body": "…ix, AVX2 fused hpel\n\nHighlights since 0.3.0 (H-30..H-34, all bitstream/output byte-identical):\n- decoder: MC call coalescing (equal-motion rects, 9.2x fewer kernel entries),\n  Arc-shared DPB (no per-slice plane deep-clones), B-slice deblock tile,\n  zero-residual recon fast path, pad-once references\n[…]\n\n  builder (build 1.66x).\n- instruments: decoder branch-level profiler stages, thread_bench and\n  decode_prof examples, XB_MBTREE lookahead arm.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "release: 0.4.0 — decoder 2.6x on real-world streams, threading race f…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T04:57:47Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "ef895dfec74c117c13ec1adb3cb9daab27c9a0c3",
          "body": "Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs: full re-benchmark snapshot — where we live after H-31..H-34",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T03:33:49Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "7f5f80f508389af5a47cb4736c8c043ffe4bff7a",
          "body": "…deblock pruned at vendor ceiling\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs(whys): H-34 — MC pad-once landed, CABAC windowed refill landed, …",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T03:26:51Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "563b71ab9b198ce64d196a1d43bd421895c63ed1",
          "body": "…m, byte-identical\n\nH-34 target 2 (CABAC bins). The engine pulled ONE bit per read_bit call — a\nbounds check, byte index and shift per renormalization shift. The reader now\nkeeps an MSB-aligned 64-bit window (8-byte big-endian refill on the fast\npath, per-byte zero-fill tail preserving the fuzzer's \n[…]\nll\nsuite + fuzz green. Residual-parse bucket ~1.05x paired (box noisy); the\nwindow also serves every mvd/skip/terminate bin outside that bucket.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "decoder(cabac): windowed bit refill — bulk 8-byte loads + lzcnt renor…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T03:26:10Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "991350ef0452b99b098abd03119373c18fd4fe6f",
          "body": "…atch\n\nThe fuzz gate caught the padded-MC brick indexing OOB when a mutated stream\nhands a reference whose plane no longer matches the caller's geometry\n(mid-stream SPS change). Fast paths now require an intact buffer (one compare\non the hot path); the extreme-MV fallback reads checked (grey beats a panic\non garbage input). Real-stream output unchanged (YUV cmp still identical);\nfull fuzz suite green again.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "common(mc): harden mc_*_padded against malformed-stream geometry mism…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T03:18:38Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "9da86f80861217c190f62da1125bb4a22a0a5c7f",
          "body": "…place, ~1.08x, byte-identical\n\nH-34 target 1 (MC filter work). Every sub-pel MC call extracted a clamped\n(bw+5)x(bh+5) halo tile before filtering (~400 B copied per call, ~100 MB of\ntile traffic per 120f real-world clip); openh264 pads the picture ONCE and\nfilters in place. RefFrame now stores edge\n[…]\ntical on x264-veryfast and own streams; decoder\nsuite green; fair stock ABBA 5/5 wins on vf.264 (median ~1.08x) and 3/3 on\nown streams (~1.07x).\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "decoder(mc): pad reference planes once (ExpandPicture) — MC reads in …",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T03:12:26Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "6e68481d9a25767db36d3e40a1d8d871f39371e2",
          "body": "…, blend re-refutation\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs(whys): H-33 — three iterations: B-tile deblock, zero-block recon…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T02:57:53Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "30f7457062e38b10a7f1632fea1b2d45cc5980ef",
          "body": "…eal streams, byte-identical\n\nTwo bricks from H-33's three-iteration descent into the remaining per-MB glue:\n\n1. deblock: the per-MB tile arm (register-resident bs derivation) was gated\n   P-only, so every B frame of a real stream fell back to the strided\n   per-edge path. bs1_tile now mirrors inter\n[…]\n\n\nGates: YUV byte-identical both streams; suites green; fair stock-vs-stock\nABBA 5/5 wins on vf.264 (median ~1.05x), flat on own P-only streams.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "decoder: B-slice deblock tile + zero-block recon fast path — ~5% on r…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T02:57:12Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "785b05f967c82de0ce614c882c6986e9eb80986a",
          "body": "…Arc'd (~1.3x)\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs(whys): H-32 — residue named, per-slice DPB deep-clone found and …",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T02:39:36Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "09d13f9aa19a8f58963a052d16562808d452b767",
          "body": "…p-clones, ~1.3x on B streams\n\nH-32 branch taps (DecMbP/DecMbB/DecBDirect/DecBMc, new INFO stages) named the\nper-MB residue: the B branch holds 40% of real-stream decode, and the ref-list\nbuilders were DEEP-CLONING every DPB entry's luma/chroma planes + motion grids\nper slice (twice for B slices: bo\n[…]\nvf.264 (with the\nMC coalescing already banked, vf decode reached 120ms this session vs 264ms\nat the morning baseline), own P-only streams +3-8%.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "decoder(dpb): Arc-share reference pictures — ends per-slice plane dee…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T02:34:56Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "6b414156768d213a1d1fbb7493e7cc963d1c34c4",
          "body": "…exploration table\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs: side-by-side SHIFT section — decoder 1.9x re-measured, updated …",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T02:21:47Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "8aa77b918592624c5fa8b3eca5ace6797ef2d2a3",
          "body": "…lookahead-only arm\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs(whys): H-31 — six side-by-side targets dispositioned; XB_MBTREE …",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T02:03:34Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "4e2c8258285bf7ad52091c6cf91a0e7579acdd18",
          "body": "…entical\n\nThe side-by-side measured our hpel build at 8x x264's hpel-filter, whose\nfused single-pass shape our scalar build_hpel_fused already mirrors (kept\nin-tree for exactly this kernel, default-off because the asm tile walk beat\nit). This is the AVX2 twin: one pass per row produces H, V and C to\n[…]\nme-hpel-BUILD bucket 52.9 -> 31.8 ms paired same-box (filter portion\n~2.7x; the remainder is plane-alloc zeroing + pad copy, a named follow-up).\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "accel(hpel): AVX2 fused half-pel plane builder — build 1.66x, byte-id…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T01:58:26Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "a54195a5a09d3603c342d352d0fc95690177632e",
          "body": "…on x264 streams, byte-identical\n\nThe side-by-side profile showed 2.41M MC kernel entries decoding a 120-frame\nx264 veryfast stream. Two per-4x4 loops were the cause: the CABAC P-inter\nrecon paid 16 mc_luma(4x4) + 32 mc_chroma(2x2) per MB regardless of\npartitioning, and spatial B-direct paid one bi-\n[…]\nl parse measured only ~3% -- the 'CABAC residue' from the side-by-side\nwas in fact this MC call storm, so dec targets #1 and #2 were one target.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "decoder(mc): coalesce equal-motion blocks into wide MC calls — 1.86x …",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T01:43:09Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "d76fbb7794363cb922de4b6233400ab0957057dd",
          "body": "Matched-settings rdtsc profiles both sides (our stage profiler + the\ninstrumented x264-prof twin), shares scaled to stock walls. Encoder: sub-pel\n167ms = 2.5x x264's entire ME; hpel build 8x; entropy emit 3.2x; we WIN\nintra-cost and pay zero lookahead (they pay 23%). Decoder: new decode_prof\nexample\n[…]\n\n2.41M times (10.4x our own streams) and ~half the decode is unattributed\nresidue = the untapped CABAC parse. Ranked exploration table included.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs: function-level side-by-side vs x264 veryfast (enc) / ffmpeg (dec)",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T01:20:17Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "a0ea82e7e1fd824343158d2aab8a367f8116576b",
          "body": "… dispatch race\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs(whys): H-30 — threading verified byte-identical after fixing the…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T01:02:36Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "16e9327197fe3f340c2fd18a43a7065e04b2dac5",
          "body": "The per-frame mv-cost SHAPE routing decision (H-24, mv_smooth mode 1) lived\nin a process-global AtomicBool. Sequential encoding is one frame at a time so\nthe global was safe; encode_all runs GOP workers concurrently and their\nper-frame stores raced — quality-preset parallel output diverged from\nsequ\n[…]\nith 4 GOPs on the quality preset, and a\nthread_bench example that asserts identity while measuring the speedup\n(~3.5-4x at 4+ GOPs on this box).\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(threads): fix mv-cost dispatch race under GOP-parallel encode",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T01:01:47Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "3a2643725cc1cc8b7bcf7ef15241e8337559e87c",
          "body": "…eriments); THE RE-BASELINE\n\n(1) Two emit-stat deployments of the mvd-distribution signal built and\nREFUTED: GOP-cumulative mis-routes early frames (fresh-IDR predictor bias,\nforeman +0.42); per-frame window flip-flops straddling content (+0.43).\nWith H-24/26/27: five experiments prove frame-granula\n[…]\n\nsuperfast and -1.6% vs veryfast (we out-compress it at 0.30x its speed).\nThe banked byte-identical bricks weigh in end-to-end. Function CLOSED.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): H-29 — signal grain law (clip-valid/frame-invalid, 5 exp…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T00:30:14Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "69f1c13596ea432a18e99664d7ff823c317858d8",
          "body": "…l three missions dispositioned\n\nRenorm averages <1 iteration/bin (683k bits / ~720k bins = 0.95 shifts/bin),\nso the clz-batch saves loop-entry mechanics only (~3-5ns/bin = 1-1.5% of\nencode) — under every floor, and the path already holds two flat bricks\n(H-16 packed output, H-18 ctx array). Two-fla\n[…]\npath CLOSED as a campaign front.\n\nOpen fronts: lookahead-deployed mvd-distribution signal (T~5 calibration\nrecorded), calm-box wall re-baseline.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs: H-28 — byte-wise CABAC engine re-priced (1-1.5%) and PRUNED; al…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T00:20:58Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "2c46cd16a35c3cbebc4bc1d2cfec9d2e6d2996be",
          "body": "… deployment REFUTED, reverted\n\nSignal: mean |mvd| among non-zero components separates every mv-cost verdict\n(foreman 4.13/akiyo 1.51 OFF vs bus 6.72/football 8.61 ON; T~5) — the axis\nmgain provably could not resolve. Deployment via within-frame learning\n(rdskip shape, freeze at 192 components) INVE\n[…]\nh measurement: the ~1-bit scalar bias\n(mode 3, RFF_MVCOST_BIAS) IS the term; richer variants unresolvable within\nthis corpus's +/-0.2 fit-noise.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): H-27 — mvd-distribution signal validated offline; online…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T00:19:53Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "12dbfb0fd2dbcae079bf3e5fc8e4fde0ab9b5b26",
          "body": "… dispatch hardened\n\nMode-range bug found by its own impossible result (identical bias tables):\nmode 3 had silently run as the dispatch, including H-25's verdict. Fixed;\ntruth's real test: bus -1.46, football -0.07, foreman +0.24 — the coherence\ninference survived. Bias b=1.0 takes foreman to noise-\n[…]\nthreshold-inseparable on mgain (0.061 vs\n0.059) — the next signal must come from the mvd distribution; the harvest\ninstrument for it now exists.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): H-26 — coherence bias explored (research modes); shipped…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-30T00:07:14Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "c568294270441696f8900a27a9987f73e3e3c221",
          "body": "…used to MV-field coherence\n\nHarvest instrument: real CABAC bits per |mvd| from the production emitter\n(BA_MVDTAB), 5-clip merge. Truth: d=0 1.11 / d=1 2.85 / d=4 5.83 / d=16\n7.49. Both analytic models wrong in opposite directions (step: mid-range\noverpriced ~1.2 bits; smooth: leave-the-predictor 3.\n[…]\naded neighbour\npredictors), which no per-vector table can express. The fix is a search-side\ncoherence term; dispatch stays the shipping default.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): H-25 — measured mvd cost table (mode 3); foreman root-ca…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T23:52:27Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "3fb1f9e29dc6ec9b6813b3163c4e1399ec60b472",
          "body": "…ver minimax)\n\nRFF_MVCOST unset -> mode 1 (dispatched by the mgain probe). Known, bounded\ncost on foreman-class content (+0.16-0.18%); bus -1.31, football -0.24,\nakiyo/mobile exact 0.00. Escape hatch RFF_MVCOST=0 reproduces the pre-H-23\nbytes exactly (verified 92626b659cde023f); new default quality hash\nd656214626e60fa8; conf_ffmpeg 8/8 pixel-exact; fast/balanced untouched.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): mv-cost dispatch DEFAULT-ON (owner's call: mean -0.27% o…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T23:47:00Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "2617c96b6f793cde345fe9e77d96b01596f17ed7",
          "body": "…man is the boundary)\n\nRouted the smooth-vs-step mv-cost shape on the same per-frame b2_mgain probe\nthat routes B2 (set_mv_smooth_mode / RFF_MVCOST[_T]); default 0 = step =\nbyte-identical, verified.\n\nDispatched T=0.10: bus -1.31 and football -0.24 keep their full wins; akiyo\nand mobile route OFF to \n[…]\nt\nattempt needs a signal from the mvd DISTRIBUTION (lookahead predicted-|mvd|\nhistogram), which measures the exact quantity the shape trades on.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): H-24 — mv-cost dispatch on the mgain probe (opt-in; fore…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T19:25:24Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "bc597432fb9cf874cd9f07d5fc231f00dc896718",
          "body": "…th mv cost model (opt-in)\n\n1. DECOMPOSE: the tail is 20,447 SIGN bits (one per non-zero mvd component)\n   vs only 11,652 bits of genuine large-vector EG3 suffix. With the ~78k of\n   context prefix those same components drag, ~14% of the bitstream is\n   'our vector differed from its predictor'. The \n[…]\nck.\n\nAlso fixed a OnceLock env-cache trap in the harness knob (first read wins,\nset_var after is inert) — it had produced a full table of +0.00.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): H-23 — mvd bypass tail root-caused in 3 iterations; smoo…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T19:16:57Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "b604394d22c0710c756ebb5b5cbd658a7d233bbc",
          "body": "…lit (29% of motion bits), intra sub-bucket\n\n(a) cb_cbp audited bin-by-bin against spec 9.3.3.1.1.4: every (A,B) neighbour\npair correct, unavailable/skip handling correct, corroborated by conf_ffmpeg\npixel-exact. The 5.7% is inherent element cost, not mis-modelling. Refuted as\na defect — do not re-o\n[…]\na_body_cabac is not yet placed, so the printed non-residual\nshare still includes intra residual. H-21's corrected ~27.7% vs x264's 21.5%\nstands.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder: H-22 — cbp ctxIdxInc audit (CORRECT, refuted), mvd bypass sp…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T18:52:33Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "fb3ed88684b4fe7b239598defaec7a11ab0679b2",
          "body": "…like syntax comparison\n\nInstrumented x264's own i_mv/i_tex/i_misc totals (guarded so the stock\nthroughput binary stays untapped). Matched run: x264 veryfast = MOTION 21.5 /\nTEXTURE 75.7 / MISC 2.8.\n\nKey: x264's 'mv_bits' is ALL non-residual MB syntax, not just motion. Mapped\nto that definition, our\n[…]\next: audit cb_cbp ctxIdxInc (5.7% on a 4-bin element is the outlier), the\nmvd context/bypass split, and split intra-body into modes vs residual.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs+ref: H-21 — x264 bit probe (X264_PROF-guarded) and the like-for-…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T18:45:16Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "8b4b30d0f04fa6c0c774c1373fa92c1ed701698e",
          "body": "…als we BEAT veryfast\n\n(1) I-slice/terminate taps: reconciliation 82.8% -> 99.7% (residue 1742 bits\n= headers+NAL+flush). Full split incl. intra MB body 17.3% of bits on 1.7%\nof MBs.\n\n(2) AQ priced on 4-QP BD (4 clips): AQ costs 1.8-5.5% BD-PSNR and earns\n7-16% BD-SSIM — its rate side had never been\n[…]\nvs superfast\n-10.5% -> -15.4%. We beat veryfast on compression. Standing rule adopted:\nquote BD with its metric AND both encoders' tuning state.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder: H-20 — I-slice taps (99.7% reconcile) + AQ rate pricing reve…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T18:38:34Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "a17c3028bcc40c89b522680d05d27a9c0c0ce1b4",
          "body": "bitacct.rs + CabacEncoder::pos(): per-syntax-element buckets as exact deltas\nof the coder's emitted-bit position, so they reconcile against the real\npayload. Observe-only, gated (RFF_BITACCT/set_enabled), byte-identical off.\nHarness: examples/bit_accountant.rs.\n\nFirst read (foreman qp27 quality, 910\n[…]\np I-slices;\nprice AQ signalling on a 4-QP BD (AQ was validated on SSIM, never on rate);\ncompare cbp/mb_type contexts vs x264 at a matched point.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder: THE BIT ACCOUNTANT (analyzer instrument #6) + first read",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T18:32:27Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "29bf6b4dc6a77c3b5e4007f07702589e6cb62333",
          "body": "…line blocked; enc-inter-code closed\n\nCABAC brick 1: the 460 context models moved from a heap Vec (pointer load +\nbounds check per bin, indexed up to 3x per encode_decision) to an inline\n[(u8,u8); 460] with a single slot borrow for read+write-back. Byte-identical\n(hashes, round-trips, suites). The w\n[…]\node: verified ALREADY decomposed by the H-6 taps — pred-buf 58%\n(traps fixed H-17), T/Q 19% (parity, H-6), recon 16%, residue 7%. No brick\nowed.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(cabac): H-18 — inline context array (byte-identical); re-base…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T17:50:59Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "512a4d28374d9cd409d0589a96976588b4fc5118",
          "body": "…vers)\n\nNon-16x16 partitions re-strided MC output per-pixel into pred_y/c_pred —\nthe runtime-width codegen trap (bounds-checked store per pixel), now\nconst-width row copies at all four sites incl. the default plan_inter_mb\npath. Byte-identical (all hashes), suites green; box too degraded for a wall\nnumber — kept as strictly-less-work per the skip-MC precedent. The stage's\nremainder is real planning work; no copy left.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(mb): H-17 — pred-buf re-stride traps fixed (4 sites, both dri…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T17:37:58Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "8d4f8ebbc8d444d35897be12d50ef7d533a47b08",
          "body": "…mplification); the coder core named as the real cost\n\nThe bit-per-byte output Vec + repack was the structural suspect; rewrote as\na packed byte accumulator (byte-exact by construction, all hashes + round\ntrips green). Measured FLAT at wall and stage — fifth confirmation that\nstreamed data movement \n[…]\n the\nx264-shape byte-wise core (clz renorm, deferred-carry byte buffer, batched\nbypass) — ~2-3x on the stage, 3-5% of encode, spec-exact output.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(cabac): packed emit output (byte-identical, FLAT — kept as si…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T17:24:29Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "bc745ef2244a96d272e7ba4539bcbf8a3b536e82",
          "body": "…ue chain collected\n\nOne geometry validation per search (plane lens, candidate window with the\nquarter +1 slack), kernel fn pointers chosen once (shape x ISA), per-eval =\ninteger bounds + phase pick + one mul + raw kernel. Full/half phases in-place\nvia Wels FFI (x2 identity), quarter via the fused s\n[…]\n34x (9/10,\nz=+2.5), mobile 1.088x (9/10), foreman 1.033 median in wild rounds — the\nH-14 ~23ns/eval prediction landed at 9-13% where resolvable.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "accel+encoder(me): MeCtx — per-search eval context; the ~23ns/eval gl…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T17:06:11Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "e55d71bddc56d7407c361707ac59f6f7c3f063b6",
          "body": "… REVISED (~23ns/eval, one chain)\n\nR1: lazy flat-variance (me_fast short-circuits it dead) + rescue dist by\nsubtraction instead of an extra SATD — byte-identical, proven vs stash-built\npre-brick binary on foreman AND bus. Wall flat because me_wide's dispatcher\nalready routes the gate off on foreman-\n[…]\nbid(unsafe)), per-eval integer bounds +\nraw phase pick + direct kernel. Prize ~15ms/24f (~12-15% of encode), the\nlargest named lever left in ME.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): H-14 — rescue-gate redundancies landed; the glue verdict…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T16:55:05Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "7fa540d367ca4b711d1ae6cae807769414822c38",
          "body": "…rung); glue residue closed at the safe-Rust floor\n\nSplit dispatch (per-frame mgain routing of 16x8/8x16 searches): structurally\nclean (foreman/bus byte-identical by construction, akiyo routed off + faster)\nbut the BD gate killed the premise — akiyo +2.45/akiyo_qcif +2.02/FourPeople\n+2.00% for 1.10-\n[…]\n-18ns kernel,\nnone clearing the brick floor; the collectible part was collected across\nA2/A3/x4/FC/B2.1). ME per-eval is at its safe-Rust floor.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): H-13 — split dispatch built and REFUTED as free (opt-in …",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T16:36:23Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "104948d17c3e49f09d2636736557350cabce079e",
          "body": "… rejected by measurement)\n\nQuality at superfast's partition shape: fair-run 1.81x faster than default\nquality, 0.28x/0.44x of superfast/veryfast, STILL -0.9% BD vs superfast on\nforeman. The subme2+SAD-force effort composition was measured and REJECTED\n(no speed over shape-only, +1.9..+8.4% BD) - kn\n[…]\nnual use.\nBus-class content pays at this rung -> the per-frame split dispatch (H-11\nbrick b) is the no-tax path. Env twin: RFF_SPLIT_T=10000000.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder: set_turbo — the superfast-class rung (shape-only, effort cut…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T16:29:51Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "b6a85b7e1ed36aeb8b711d67358d41c0f0605b68",
          "body": "…ons x ~1.8x effort x ~1.8x glue)\n\nIso-shape experiment (RFF_SPLIT_T=inf): P16x16-only quality runs 1.81x faster\nand STILL beats superfast by -0.9% BD. The 'murder' is two priced policies\nstacked on a ~1.8x glue residue — not a missed function.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs: H-11 six-whys — the 5.9x vs superfast decomposed (1.81x partiti…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T16:23:18Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "132ad54da8f83b0e1ba47768671dfda71f754a0b",
          "body": "One level maps onto (ring pattern x B3 iteration budget): 5=ring8 uncapped\n(quality default, unchanged), 4=cap3, 3=cap2, 2=ring8 single pass\n(= balanced), 1=ring4 single pass. AB_SUBME prices the family.\n\n5-clip 4-QP BD + deterministic foreman work column: subme3 is the star —\n1.92x less sub-pel wor\n[…]\nnded dial between\nbalanced and quality. subme4 recorded as dominated by subme3. Defaults and\nhashes unchanged (subme5 == today's knobs exactly).\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): set_subme — the priced sub-pel effort ladder (H-10)",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T16:09:19Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "2d9fdd8f0ff3df3728103636a96ee74d9664b386",
          "body": "…e rings\n\n8-wide x4 batching removed from the hot arms (per-candidate evals inside the\nSAME argmin — value-identical, both hashes proven unchanged): the 8-wide\nkernels lose to per-candidate Wels asm. The remaining regression named the\nreal mechanism: an 8-point argmin pass pays 8 evals per re-centre\n[…]\ns\nfully monotone (worst crew SSIM +0.03). Speed verdicts at 12 paired rounds\nare outside any noise floor — this is a resolution, not a deferral.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): SP-FC flip CLOSED as a prune — fixed-centre inflates wid…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T16:02:44Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "7ec496bc9d15f3704b88e828945c7aa1977f67cc",
          "body": "sad_x4 / satd_x4 / satd_x4p / satd_avg_x4 generalized: h-parametrized 16-wide\ncores + new 8-wide cores ([row r|row r+4] lane packing; SAD pairs consecutive\nrows — order-free). x4_shape() gates every wrapper. The FC diamond (default)\nand FC ring (RFF_SP_FC) now batch every ME partition shape.\n\nGates:\n[…]\n,\nfootball -0.98 (was -0.52), crew -0.29, mobile -0.06, foreman +0.13 (the\nfamily's known +/-0.15 fit-noise tail). Mean -0.69 — kept default-on.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "accel+encoder(me): x4 kernel family bridged to 16x8/8x16/8x8",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T15:40:17Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "6a9c42e7464139b6017d1cff1c0453f39b581012",
          "body": "…lock\n\nMinor bump, not patch: encoding with stock settings now emits a DIFFERENT\nbitstream than 0.2.x with no code change on the caller's side. CABAC is\ndefault-on, which moves the profile default from Constrained Baseline to Main,\nand adaptive quantization is default-on. RUSTY_H264_LEGACY_CAVLC=1 r\n[…]\nrected.\n\nCHANGELOG [Unreleased] promoted to [0.3.0] with the changed-defaults warning\nfirst. README install snippets moved to \"0.3\".\n\nCo-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "release: 0.3.0 — changed encoder defaults, and break the publish dead…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T15:31:20Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "83dea8f886858b38fe9cb17ba688a59ca5308e56",
          "body": "…ing (opt-in)\n\nSub-pel anatomy vs x264-veryfast: per-eval KERNEL PARITY; the whole loss is\nstructure (~24 evals/MV ring8-converge-2steps vs their ~8-9). Blanket cuts\nstay BD-refuted; the lever is their fixed-centre batched evaluation.\n\nsatd_avg_16x16_x4 (four fused avg+SATDs, shared source band) + t\n[…]\nign, improved from +0.18 half-only). Speed: box at +/-30%/round this\nsession — no honest verdict; ships OPT-IN pending one quiet-box paired run.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): satd_avg_16x16_x4 — the completed fixed-centre sub-pel r…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T15:28:11Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "b3cf69083b83a81672df85dfb27edcb8f3058e28",
          "body": "The crate READMEs added in the previous commit described a couple of knobs\nwrongly, and the root README + CHANGELOG still described the encoder as it\nstood at 0.2.1 — before CABAC encoding, B-frames, AQ, mb-tree and the 8x8\ntransform landed. Both fixed against the source, not from memory.\n\nCorrectio\n[…]\nonly. Workspace builds, doctests pass. Staged by path — the concurrent\nsatd_avg_16x16_x4 work in the tree is deliberately untouched.\n\nCo-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs: root README + CHANGELOG catch-up; correct the crate profiles",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T15:24:28Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "d083d864f2347cc34b231817702056eb97d48514",
          "body": "…t-in), T/Q attribution closed\n\n1. Re-profile: sub-pel still ~40% of encode; caught the fused hpel builder\n   running 2.2x SLOWER than the asm-kernel tile walk (950 vs 617 us/frame) —\n   default flipped back to tiles (RFF_HPEL_FUSED=1 keeps the fused base for a\n   future AVX2 kernel). Revert-if-not-\n[…]\nates: suites green; default-off byte-identity held per knob; full-restore\nanchor (RFF_ME_SADFP=0 RFF_ME_FC=0) still reproduces e11235654539ba44.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): goal session — re-profile, residue named, sub-pel FC (op…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T15:17:19Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "1e2d4ae2449c1324222d38d0cf55beddab532350",
          "body": "Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs: lets-win-optimize status ledger after the Track-B arc",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T15:03:02Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "49e66e0ce3db5cff1cfde7ef518624e9e8cc22bd",
          "body": "…s hpel builder\n\n- satd_16x16_x4 (custom AVX2, shared-source-band Hadamard batches): fixed-\n  centre argmin diamond now covers SATD-routed frames too. Own BD gate at\n  mode 0 (AB_FC/set_me_fc): monotone non-regression — bus -1.93, football\n  -0.52, crew -0.22, foreman -0.11; worst +0.06 SSIM (noise)\n[…]\nGates: oracles exact; suite green; conf_ffmpeg 12/12 pixel-exact; paired\n  wall vs pre-B2 anchor: bus 1.108x (7/8), foreman 1.031, mobile 0.994.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me)+common: FC everywhere via satd_16x16_x4; fused single-pas…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T14:57:23Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "52399d300cba45bb028d777dbd3aec793b30d21d",
          "body": "…16x16_x4 batch kernel\n\nThe SADT sweep refuted its own premise informatively: at T=0.155 soccer read\nWORSE (+0.18 vs +0.09) with FEWER frames routed ON — non-monotone, so the\n+0.09 tail is BD-fit noise, not regression. T stays 0.13 and the flip bar is\ncleared: RFF_ME_SADFP unset now means mode 1 (di\n[…]\nl == GOP-parallel.\nvs x264 (quality): foreman -10.6% superfast / +3.9% veryfast (session start\n-9.9/+4.6); bus -1.4% superfast / +7.0% veryfast.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): flip B2 dispatch DEFAULT-ON; fixed-centre diamond + sad_…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T14:43:19Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "1932e24f59d6d2c62bc7ae6947d465fe085a58b1",
          "body": "…y corpus loss zeroed\n\nPer-frame content dispatch for B2 (SAD full-pel), built on the R6 discipline:\nfive candidate signals instrumented offline against the 16-clip truth table\n(examples/b2_signals.rs) before choosing the axis.\n\n- Axis: mgain = (SAD@0MV - best over +/-8 step-4 grid)/SAD@0MV — the\n  \n[…]\nale-binary strike four: the first dispatched corpus run reproduced the\nforce-on table digit for digit — me_ablation.exe predated the dispatcher.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): the B2 dispatcher — mgain axis + dcfrac flash veto; ever…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T14:29:08Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "32d2490f2e82748823d1a5374136d0beeb534742",
          "body": "…a content-dispatched tool\n\nRFF_SP_MAXIT / set_sp_maxit: cap each sub-pel step at N ring passes (0 =\nunlimited = byte-identical, hash-verified) — the bounded budget x264 subme\nlevels have always had, built to test whether it unlocks B2 speed.\n\nPriced (AB_SPCAP, 4-QP BD, 6 clips + paired ABBA):\n- cap\n[…]\nrd discounts). Next brick: instrument 3+\ncandidate signals (frame mean-luma delta first) against the 16-clip truth\ntable per the R6 methodology.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): B3 sub-pel iteration budget (opt-in) — the B2 family is …",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T14:05:30Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "d633e2e9dcb72df05599a4ae1c434e5a4ebe545f",
          "body": "…raming, Tracks A+B2)\n\nThe instrumented per-call finding (ours 39,996 searches @ 1.68us vs x264\n53,527 @ 0.16us) retires the candidate-count framing; the plan maps every\nstage of the 3.8x gap to campaigns, and Descents G-H log Track A (landed,\nbyte-identical) and Track B2 (built, calibrated, opt-in on a sign-flip).\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs: lets-win-optimize battle plan + WHYS Descents G-H (per-call ref…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T13:59:57Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "897dca69c3f8f08414c63f75c6e48b2a0ec42b4d",
          "body": "…ch hoists, SAD full-pel (opt-in)\n\nTrack A (byte-identical, all six clip x preset hashes unchanged):\n- A2: mc_satd_hp — the per-candidate cost closure re-derived the plane-cache\n  OnceLock (twice on the quarter-pel arm), the RFF_HPEL_REF knob, and the\n  source slice base for EVERY candidate; they ar\n[…]\nx264 (foreman, quality, B2 on): -10.5% BD vs superfast (was -9.9),\n+4.0% vs veryfast (was +4.6).\n\nFull log: docs/WHYS-speed-gap.md Descents G-H.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): Challenge-1 Tracks A+B — fused avg+SATD kernel, per-sear…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T13:59:57Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "6a7209045f650cd47b2552d48c3d9bdfeaff565e",
          "body": "…nements; README refresh\n\nPre-session working-tree state from the instrumented function-by-function\ncomparison (video-tests/x264_instrument.py as the x264-side twin of prof.rs):\nencode_hash gains EH_GOP so both encoders' stage breakdowns run at matched\nkeyint; x264_bdrate refinements from the same pass; README updated.\n\nCo-Authored-By: Claude Fable 5 <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "bench: EH_GOP knob for reference-matched breakdowns; x264_bdrate refi…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-29T13:59:32Z",
          "body_truncated": false,
          "is_coding_agent": true
        },
        {
          "oid": "fbc8cb4ea6a99fa82356b71a7a857f4ab1e36f3a",
          "body": "…acked\n\nThe bench/_map tree is decoded YUV, frame dumps and probe binaries produced by\nthe RD harness: 831 files, ~31 GB on disk, and 4.16 GiB had been committed\nacross 67 blobs. Five exceeded 137 MB, above GitHub's 100 MB hard per-file\nlimit, so the branch was unpushable.\n\nStripped from the 52 unpu\n[…]\nnpushed range — commits already on origin keep their SHAs) and ignored here\nso it cannot recur. The files remain on disk, untracked.\n\nCo-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "chore: gitignore bench/_map — generated RD artifacts must never be tr…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-27T23:08:13Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "a76d9a41ec9796d15bc5627e6d701fc75f7fb6b0",
          "body": "…bility mode\n\nCABAC/Main is now the shipped default, so the fair x264 anchor is\n`--profile main` rather than baseline — the old XB_CABAC opt-in no longer\ndescribes what we ship.\n\nXB_ALLTOOLS additionally enables every tool we implement but leave off by\ndefault (8x8 transform, multi-ref P, mb-tree, a\n[…]\n\n\"gap from tools we lack\" from \"gap from efficiency in the tools we have\",\nwhich the shipped-default comparison alone cannot answer.\n\nCo-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "bench(x264_bdrate): re-anchor to --profile main; add XB_ALLTOOLS capa…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-27T22:50:52Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "77a2bb6c0404fc8f7e9f0dbd77cef5b8eccbfd50",
          "body": "…to 32\n\nTwo byte-identical bricks from running the three-iteration process over both\nhalves of the mc_luma split (search-fallback 76-79%, recon/skip 21-24%).\n\nTRACK A -- recon + skip-check via the plane cache.\nTagging resolved the remaining population: recon 41-67%, skip-check 24-35% (and\nskip-check\n[…]\n probe at the\nlevel above the change.\n\nGates: byte-identical on 8 clips x balanced/quality; 113/113 suite; 28/28 ffmpeg\npixel-exact.\n\nCo-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(mc): route recon/skip through the plane cache, widen the pad …",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-27T22:44:29Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "8cda0721a69469398da72d52d640b08e7f29ab84",
          "body": "… byte-identical\n\nRe-probe of a prune Descent B made on a single measurement. The three-probe rule\nexists because a wrong refutation is permanent: nothing revisits it.\n\nThe prune was priced on the wrong denominator TWICE.\n\n(1) Calls, not time. A full-pel 16x16 is a row copy; a quarter-pel is a\nper-p\n[…]\n byte-identical on 8 clips x balanced/quality with the two binaries proven\ndistinct by md5; 113/113 suite; 24/24 ffmpeg pixel-exact.\n\nCo-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): serve edge full-pel from the padded plane — up to 1.28x,…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-27T22:13:32Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "d8de9134e21ba857e894d1c9c465fda8b06d6374",
          "body": "…entical\n\nSub-pel refinement is 41% of encode after the diamond-ladder and half-pel\nin-place changes. It walks an 8-point ring around a MOVING centre, so\niteration N+1's ring necessarily re-contains the previous centre and several\nprevious ring points. Census: 43.5% (mobile) / 43.4% (football) / 40.\n[…]\nnt rate is NOT evidence of low value.\n\nGates: byte-identical on 6 clips x balanced/quality; 113/113 suite; 24/24\nffmpeg pixel-exact.\n\nCo-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): memoize the sub-pel ring — 27% fewer cost evals, byte-id…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-27T21:50:16Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "f130903bb6f4e0362b58b3f4df67819857c92dae",
          "body": "WHYS-speed-gap.md is the persistent descent log -- S1-S6, R1-R11, U1-U6,\nU5-struct, and Descents A-C -- recording every measurement INCLUDING the\nrefuted hypotheses, so a later session does not re-run a dead branch. Refutations\nlogged: the sub-pel skip-gate, the ring-1 online dispatcher, the\nlambda-\n[…]\nnter-gap.md, me-subpel-and-snap.md, rd-skip-adaptive.md and\ndeblock-bs-precompute-plan.md carry the same discipline for their areas.\n\nCo-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "docs: six-whys descent logs for the speed and inter-coding gaps",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-27T21:31:05Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "2c1fe86bba7c7ba67e6b6d50e54a14dd0917563b",
          "body": "fetch_clips.sh + manifest.tsv reproduce the 20-clip corpus (CIF through 1080p,\nstatic through fast-pan) that every BD-rate verdict in this campaign is gated\non. A single clip misleads in both directions -- the diamond-ladder result was\nnegative on 18 of 20 clips but +6.06% on bus_cif for one candida\n[…]\ns reproducible without putting multi-GB source video in git.\n\nanalyzer/ is the per-stage/per-primitive comparison tool against x264.\n\nCo-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "video-tests: real-clip speed/quality corpus harness",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-27T21:31:05Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "ed591253d9871fd832056f351da83e0b4321579b",
          "body": "The descents are only as good as their instruments; these are the ones that\nproduced (and repeatedly corrected) the numbers.\n\nconf_ffmpeg      -- STRICT external conformance. Encodes a real clip and\n                    requires ffmpeg's decoder and ours to be PIXEL-IDENTICAL.\n                    Our\n[…]\np numbers unchanged); and a profile-feature build\ninflates wall time ~1.6x, which once read as a 2.6x regression that did not\nexist.\n\nCo-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "bench: measurement instruments for the ME speed/quality campaign",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-27T21:31:04Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "61f038b2c8bd3c56e5b2ba0e3566dfe081b3a061",
          "body": "Two levers from a six-whys descent into enc-me (72% of encode), plus the\naccumulated campaign work in the same files.\n\nDESCENT A -- the coarse-to-fine ladder was buying NOISE, not motion.\nPer-rung census: the 64 and 32 rungs are ~39% of full-pel evaluations at a\n0.05-0.84% hit rate, and eval counts \n[…]\nality x\nqp24/33). Re-priced vs x264: foreman -9.2->-9.9% vs superfast and +5.5->+4.6%\nvs veryfast; mobile -3.2->-3.6% / +1.6->+1.2%.\n\nCo-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): kill the coarse diamond rungs + SATD half-pel in place",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-27T21:30:31Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "c9202a54d14eadc8f25c675eac30c90d38db62d2",
          "body": "…p (enc+dec)\n\nThe CABAC P path was hard-capped to a single reference (lib.rs `p_dpb = dpb[..1]`)\nbecause neither the encoder emit nor the decoder parse coded ref_idx under CABAC —\nso --refs N only ever helped CAVLC. This wires ref_idx_l0 CABAC on BOTH sides.\n\nEncoder (mb16.rs): `cb_ref_idx` (unary b\n[…]\n, richer than CAVLC's ~0.5%). Stays opt-in via\n--refs N (multi-ref costs N× the ME); adaptive ref-count dispatch is the follow-up.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "codec: CABAC multi-reference P — code ref_idx_l0, remove the 1-ref ca…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-23T21:27:41Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "e8053aef8f016fcc45d85613d0958848d70fbaba",
          "body": "…oo (-8% bus)\n\nAn exhaustive ±24 diagnostic revealed the integer ME leaves 6-15% BD on the table\non fast motion (bus -15%, football -11%, stefan -8%, foreman -6%): the coarse-to-\nfine diamond stalls before reaching far/fast MVs, exactly as it did on flat\nsurfaces. The me_wide rescue already fixes fl\n[…]\narns 7-8%. Default-on for Quality (me_fast, me_range env-tunable).\nPixel-exact vs ffmpeg; Fast default byte-identical; tests pass.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): fast-motion rescue — the diamond stalls on FAST motion t…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-23T18:57:27Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "252bf32c2d9d1d865acb37fd57a74091dd3c9b3b",
          "body": "… gate proven futile\n\nA rigorous discovery harvest (prom search-skip-gate template) settled sub8x8's\nlong-standing opt-in status. 12 real Derf CIF clips × 4 GOP-segments = 48 samples,\neach labelled by encode-both-ways sub8x8 marginal BD, harvested with all 6 motion\nchannels (divergence/complexity/ac\n[…]\n Fast never runs it, so the Fast default stays byte-identical. Pixel-exact\nvs ffmpeg; tests assert the Quality default enables it.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(inter): sub8x8 DEFAULT-ON for Quality — net real-content win,…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-23T16:42:07Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "94dbac83a8eb8d44fd2484b229fde57eca7fb4a5",
          "body": "…m / +4% tsrc)\n\nThe requested \"asm SATD grid kernel\", scoped by profiling. The SATD is already\nAVX2 (openh264 WelsSampleSatd16x16_avx2), and its Hadamard can't amortise across a\nsliding grid, so a from-scratch kernel has no algorithmic headroom. What IS left is\nper-call overhead: mc_satd's interior \n[…]\n asm grid kernel here: the big win was the interpolation\nfix (2.21x→1.19x), not the SATD. Fast default byte-identical; tests pass.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): batched full-pel rescue grid over the AVX2 SATD (+7% zoo…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-23T15:07:10Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "195b4d4a64bc79891cf66a852d1bc999cfde8cfe",
          "body": "Set out to add an asm SATD grid kernel for the me_wide rescue; profiling\n(RFF_MC_PATH) found the SATD is ALREADY AVX2 (openh264 WelsSampleSatd16x16_avx2\nvia satd_px) and the real cost was elsewhere: on zoom, 89% of the rescue's mc_satd\ncalls took the mc_luma INTERPOLATION path, not SATD (2.0M interp\n[…]\nhe kernel — the bottleneck was interpolation, not\nthe (already-asm) SATD; an asm SATD kernel would have optimised the wrong thing.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): snap the rescue grid to integer-pel — 2.21x→1.19x on zoom",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-23T15:03:00Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "52d898b57ec1857b33ff3744dd0226e852c16174",
          "body": "…t/fractal waste\n\nDeep-dive into rot (me_wide's worst speed/quality point: 2.25x for -0.07%) found\nthe rescue's fire signal and its payoff are DECOUPLED exactly where it hurts. The\nfine grid fires on ~90-98% of flat blocks on rot/mand but improves <12% of them by\n~1.5%, vs tsrc/zoom improving 44% by\n[…]\nx (was 2.11x)\nPixel-exact vs ffmpeg, deterministic, Fast default byte-identical, tests pass.\nRFF_ME_LEARN / RFF_ME_PAYOFF tunable.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): online per-frame rescue-payoff gate — fixes me_wide's ro…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-22T22:49:45Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "237d8755d848424a2d4a14a26853dc896bc607d4",
          "body": "…aptive)\n\nThe coherence gate made me_wide never regress, so ship it as the Quality-preset\ndefault. Validated on a 6-clip corpus (added zoom/rotation/mixed-motion to the\nsmooth/pan/fractal set) — me_wide-only vs off, CABAC + B-frames:\n\n  tsrc -10.31%  zoom -3.40%  mand -0.41%  rot -0.07%  mixed -0.02\n[…]\nid opt-in\nendpoint per the adaptive discipline; gating it is future work.\n\nTests pass scalar + accel; Fast default byte-identical.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): me_wide DEFAULT-ON for the Quality preset (proven all-ad…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-22T16:05:09Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "3bb3073326e5c90a3fbaa092aadf3782da358331",
          "body": "…h-B-frames)\n\nme_wide regressed panc +0.65% ONLY with B-frames (P-only it was neutral). Root\ncause: panc is a near-static PURE PAN — a single global MV predicts the whole\nframe, so the diamond's seed (the median = that MV) is already right and never\ngenuinely stalls. The rescue then only finds SPURI\n[…]\nn?) — invisible to the local search; compute it\nonce per frame. me_wide default-on-for-quality is now unblocked (never regresses).\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): coherence gate — me_wide never regresses (fixes panc-wit…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-22T15:40:00Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "ff9777727818daaac9e1117920b9e51ebcdb4276",
          "body": "…in at low cost\n\nReplace the variance-proxy dispatch (which fired the wide search on flat-but-well-\npredicted blocks) with a precise diamond-STALL detector, and use the fine step-2\ngrid (the -17% version) now that it fires only on genuine stalls.\n\nThe stall signal is a CONJUNCTION: a FLAT source blo\n[…]\nvar env-tunable. Pixel-exact vs ffmpeg (CAVLC+CABAC, scalar+\naccel). Default byte-identical. me_wide still opt-in (default false).\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): diamond-stalled RESCUE trigger — recover the fine-grid w…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-22T15:08:43Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "8bf9971bb7eea8331393a27301d0ca0817519976",
          "body": "…surfaces\n\nThe tsrc deep-dive (P-frames +22.6% behind x264's *simple* dia/subme1) traced to\nthe motion search: our coarse-to-fine orthogonal diamond is gradient descent, and\non a smooth clip's FLAT cost surface it stalls at a plateau near the seed, never\nreaching the better MV that exists within ±16\n[…]\nescue trigger vs the variance proxy) recovers it.\nLikely the biggest real-content lever — the diamond stalls on any smooth region.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(me): adaptive wide search — fix the diamond stalling on flat …",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-22T14:52:29Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "93123ef6002547484d18de83de017fda92b9f074",
          "body": "…tive\n\nAdd P_8x8: a P macroblock may split into four independent 8x8 motion partitions\n(mb_type 3, four sub_mb_type=0, four MVs) — finer motion granularity on complex /\nboundary motion. The #1 item on the x264-gap roadmap, and an encoder-only brick:\nthe decoder already parses P_8x8 fully (decode_p8x\n[…]\nON inefficiency (over-coded luma, not skip/RDOQ/AQ/deadzone) that\npartitions don't address — a future motion-search-quality brick.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(inter): P_8x8 sub-partition motion — conformant, content-adap…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-22T14:03:39Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "619d40bc8bc3a255a0bd37da820bbd38668ac270",
          "body": "…ks the base\n\nmb-tree previously did nothing in the bitrate (RC) path. Enable it: the encode_all\nbitrate branch runs the per-GOP lookahead, sets each frame's per-MB offset, and the\nrate controller supplies the frame base QP. The per-GOP-centered offsets are\nrate-neutral per GOP, so the controller's \n[…]\nhe controller's real operating range it is\n+0.03%. Always BD-rate an RC feature within the range the controller can actually hit.)\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(mbtree): run in rate-control mode — per-GOP centering, RC pic…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T21:52:18Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "5038859c8ff5a41e98b6bf1a4ecdf6db433c0987",
          "body": "…f, score full)\n\nPer request, default the mb-tree lookahead to the fastest half-res motion search\n(~4x cheaper lookahead), and add a Hybrid mode that recovers full-res quality.\n\n- LookaheadMode enum { FullRes, Hybrid, HalfRes } replaces the mbtree_halfres bool;\n  cfg.mbtree_lookahead (default HalfRe\n[…]\ns ffmpeg (mb-tree only moves QP). Default (mbtree off)\nbyte-identical. panc stays neutral in every mode (predictability back-off).\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(mbtree): half-res lookahead default + Hybrid mode (search hal…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T18:48:13Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "b5aee314e6c5e8a7310316027ff1adc974e7f5b6",
          "body": "…ured trade)\n\nGeneralize the mb-tree lookahead cost functions to a block-size parameter and add\nan optional 2x-downsampled (half-res) motion search: 8x8 blocks on half-res\nplanes, ~4x cheaper ME (~2.7x faster encode, 2055->768ms at gop24). MVs are scaled\nback to full-res quarter-pel so the propagati\n[…]\ns / --mbtree-halfres 1, for users who want the speed.\n\nConformant (pixel-exact vs ffmpeg — only moves QP). Default byte-identical.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(mbtree): half-resolution lookahead — opt-in speed lever (meas…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T18:04:49Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "d7d4da6c6f7a54be63273c7e011e8c9275c58623",
          "body": "mb-tree previously only fed the CAVLC CQP-P path. Thread the per-MB QP offset\nthrough the CABAC slice encoders (I/P/B) and the B-frame reorder pipeline.\n\n- Shared `apply_mbtree_qpo(aq_qp, qpo)` helper folds the offset into the spatial-AQ\n  map; used by encode_slice_data (CAVLC) and encode_slice_data\n[…]\n         CABAC-P    B-frames(2)\n  mand       -0.18%       -0.28%\n  panc       +0.00%       +0.02%\n  tsrc       -1.66%       -1.46%\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(mbtree): extend temporal AQ to the CABAC and B-frame paths",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T17:56:40Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "6d36d3f3c3e6445f20a4d248d12fbdf69fc7f0df",
          "body": "…esses\n\nMacroblock-tree adaptive quantization: the temporal complement to the spatial AQ\n(and the tier4-lookahead branch's namesake — the first real lookahead pass). A\ncheap per-GOP forward pass over the SOURCE frames propagates each MB's\nfuture-reference importance backward along motion vectors and\n[…]\nc -1.80% — a win or the no-regression floor on all content.\n\nEnv RFF_MBTREE_DBG dumps the per-GOP spread/residual_frac/eff signal.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(mbtree): lookahead temporal AQ — content-adaptive, never regr…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T17:43:21Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "a7c9d2ca63fd66c26ce3e6876aea4a2e565f5117",
          "body": "…aptive all-wins\n\nBring up the High-profile 8x8 transform on INTER macroblocks (P and B), the\nexact forward inverse of the decoder's t8x8 inter luma path. Per-MB the encoder\nRD-picks between four 4x4 DCTs and one 8x8 DCT of the motion-compensated\nresidual; transform_size_8x8_flag is written after cb\n[…]\nal) MBs clear\nthe bar; fine-texture MBs gate off to the no-regression floor.\n\nEnv knobs RFF_INTER8 (0/1) + RFF_INTER8_PEN for A/B.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(8x8): inter transform_size_8x8_flag — conformant + content-ad…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T17:00:50Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "87cd6056c5bbf0399d105f8c98c98b254d7f0a08",
          "body": "… RD choice\n\nThe full 8x8 transform intra encoder, gated pixel-exact vs ffmpeg (our decoder\nalready decodes 8x8 — it's the round-trip oracle). Every brick:\n- High-profile SPS (profile_idc 100: chroma_format_idc/bit-depths/no scaling\n  matrices) + PPS extension (transform_8x8_mode_flag). config.trans\n[…]\nld tighten it). Opt-in\n(default false); default CAVLC byte-identical (13fe5606); 92 tests green + new\ntransform_8x8_intra_decodes.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "transform(8x8): I_8x8 High-profile encoder — conformant, 3-way per-MB…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T04:06:22Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "c1d86307037b2b9cecdc347113c75242d011483e",
          "body": "…oder foundation)\n\nFirst brick for the High-profile 8x8 transform (the decoder already decodes it —\ndecode_i8x8 / inverse_quant_8x8 / un_scan_8x8 / add_residual_8x8 — so it's the\nround-trip oracle). The forward direction the ENCODER needs:\n- forward_core_8x8 + fwd_1d_8x8: promoted from #[cfg(test)] \n[…]\n 4x4 blocks,\ncoeff k of sub-block s -> 8x8 scan 4k+s). 8x8 CABAC residual is a separate later\npath (decoder has no CABAC 8x8 yet).\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "transform(8x8): expose forward 8x8 transform + add forward quant (enc…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T03:32:43Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "9d22b015b7fca427153df4d18add5d2a35638770",
          "body": "RDOQ was plumbed into the scalar quant sites; the accel (asm, x86_64 --features\nasm) I_16x16 + chroma paths still hard-quantized via quant_four_4x4, so the\ndeployed asm build only got RDOQ on I_4x4 (plan_i4x4 is always scalar) — enough\nfor high-detail I4-dominated content (mand asm==scalar) but miss\n[…]\n ffmpeg decodes it PIXEL-EXACT. Scalar path unaffected\n(cfg(accel) only); default CAVLC byte-identical (13fe5606); 91 tests green.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "cabac(rdoq): port trellis quantization to the asm (accel) path",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T03:27:57Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "98d8aee1659421f4c8b5f21ee569aa09106d6595",
          "body": "…n all-intra\n\nCABAC trellis quantization (RDOQ): each 4x4 residual coefficient's level is\nRD-optimized (min SSD + λ·R_cabac) instead of the distortion-only hard quantizer.\nTwo passes per block: (1) per-coefficient level lowering |q|->|q|-1; (2)\nlast-significant-position trimming — zeroing a trailing\n[…]\nfault-masks-config-default bug (map_or(Ok(0.0)) ->\nmap_or(Ok(cfg.cabac_rdoq))). Default CAVLC byte-identical (13fe5606); 91 green.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "cabac(rdoq): trellis quantization — structure-adaptive, -0.5..-1.3% o…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T03:16:16Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "5eddf4098addde7f09ac2d2df60cb0b01d86e356",
          "body": "…eadzone recorded\n\nApplied the adaptive workflow (measure → dispatch or prune) to the CABAC RD\ndecision, on a real BD-rate harness (ffmpeg PSNR + Bjontegaard; validated: CABAC\nvs CAVLC = -8.90% on mandelbrot CIF, identical PSNR/QP confirming the pure\nentropy win). Two experimental knobs added (defau\n[…]\na per-coefficient RDO, not a dispatch.\nKnobs kept behind safe defaults as the ledger + future real-content hooks.\n\n90 tests green.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "cabac(rd): BD-rate-measure the CABAC RD levers — two washes pruned, d…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T02:15:47Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "aa091421eaf892577c1e8aef567b1dff981260a3",
          "body": "Completes CABAC for every slice type. The B inter syntax inverts the decoder's\ndecode_slice_data_cabac B-slice path for the modes the encoder produces —\nB_Skip, B_Direct_16x16 (0), B_L0/L1/Bi_16x16 (1/2/3), all 16x16 (no sub_mb_type,\nno intra-in-B):\n- cb_mb_type_b (ctx 27, bci from neighbour mb_dire\n[…]\nALSELY read as a\ndesync — CABAC's 1-ref P vs CAVLC's 2-ref P legitimately differ; ffmpeg-vs-ours\nis the correct conformance gate.)\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "cabac(encoder): B-slice CABAC — full I+P+B CABAC now conformant",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T01:53:55Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "cd32397ca7cee9d1a6ce49fbf84311f17f9d932b",
          "body": "…ent (pruned)\n\nApplies the adaptive workflow (measure → dispatch or prune) to CABAC's natural\ncontent-dependent knob: the P/B slice header's cabac_init_idc ∈ {0,1,2} picks one\nof 3 context-init tables. Added cfg.cabac_init_idc (default 0), threaded to the P\nslice header + CabacEncoder, plus --cabac-\n[…]\ner-selectable, the foundation if a larger real-content corpus\nlater shows a real spread) with the safe default 0.\n\n89 tests green.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "cabac: content-selectable cabac_init_idc + adaptive-dispatch measurem…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T01:37:02Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "e67fddf44388d548b7b7c3c8841cbd9a25835a96",
          "body": "…al content\n\nThe CABAC inter syntax, inverting the decoder's decode_slice_data_cabac P-slice\npath: mb_skip_flag, mb_type_p (P_L0_16x16/16x8/8x16 + intra-in-P), mvd (UEG3\nbinarization via cb_ueg_mv + the 30-entry mvd/ref neighbour cache mirroring\nWelsFillCacheInterCabac), and the inter residual (cb_r\n[…]\nral, -35..-59% near-static. CI test\ncabac_p_roundtrips_and_matches_cavlc. Default CAVLC byte-identical (13fe5606);\n90 tests green.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "cabac(encoder): P-slice CABAC — conformant, -5..17% vs CAVLC on natur…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T01:31:54Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "200d94de3238cae18f32347c43c5bb6dfbf9a697",
          "body": "…it_inter_cavlc\n\nPrep for the CABAC inter backend (P/B CABAC). Mirrors the intra plan/emit\nsplit: `plan_inter_mb` does the per-partition motion compensation + residual +\nquantize + reconstruct + motion-grid commit (returning an `InterPlan`), and\n`emit_inter_cavlc` does the CAVLC entropy coding. A se\n[…]\nafter (P sha1 13fe5606, B sha1 267fdd99). encode_inter_mb_v1_b keeps its\nsignature (wrapper = plan + emit); all callers unchanged.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(refactor): split encode_inter_mb_v1_b into plan_inter_mb + em…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-21T01:18:19Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "1efda7e5436b143a82bcc92b5b39d2abc6201dd9",
          "body": "… CAVLC\n\nThe CABAC I-slice syntax layer, the exact forward inverse of the decoder's\ndecode_slice_data_cabac I-slice path. Every binarization + context selection\nmirrors a parse_*_cabac; the neighbour state (nzc cache, cbf_dc, cat, cmode,\nmb_cbp, last_delta_qp) is reconstructed identically so context\n[…]\nhe fuzz suite): a malformed luma/chroma_log2_weight_denom overflowed\n`1 << denom`; now rejected per spec [0,7] so it never panics.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "cabac(encoder): I-slice CABAC entropy coding — conformant, -7..12% vs…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-20T17:14:42Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "e2fedab73efd3c6047c3ada4117b5c32ebc17f9b",
          "body": "Prep for the CABAC entropy backend: separate the intra macroblock's\nmode-decision + transform + quantize + reconstruct (now `plan_mb`, returning\nan `MbPlan`) from its CAVLC entropy coding (now `emit_mb_cavlc`). This lets a\nsecond backend (`emit_mb_cabac`, next) reuse the ENTIRE non-entropy path\nbit-\n[…]\nCAL before/after\n(sha1 13fe5606, 66108 bytes). encode_mb keeps its signature; both call sites\n(real emit + trial_intra) unchanged.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(refactor): split encode_mb into plan_mb + emit_mb_cavlc",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-20T16:49:33Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "c91d1fffc0f74305a42947e925f2fb1006a6a656",
          "body": "…ngine\n\nFoundation brick for the CABAC encoder (largest remaining gap to x264,\n~10-15% BD-rate over CAVLC). Two moves, both zero-risk:\n\n- Move the 3 CABAC constant tables (RANGE_LPS/STATE_TRANS/CTX_INIT, 604\n  lines) from the decoder into rusty_h264-common (cp verbatim, pub), so the\n  encoder can sh\n[…]\nomoted module — every bin + terminate round-trips\nbit-exact. Syntax layer (mb_type/cbp/residual + PPS entropy_coding_mode=1) next.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "cabac(encoder): share tables via common + promote the proven encode e…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-20T16:40:42Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "cf3590dc34c8719aa1041a23a6674be0f1f766af",
          "body": "…bframes auto`\n\nWeighted bi-pred was the prerequisite: the plain (p+q+1)>>1 average is only correct\nfor the EQUIDISTANT B of a single-B gap, so bframes>1 (unequal-distance B's) LOST.\nNow the PPS signals weighted_bipred_idc=2 (only when bframes>0 — B-less PPS stays\nbyte-identical) and every bi-blend \n[…]\nignal. Busy content still → P-only; per-GOP B/P\ndispatch + QP cascades + AQ all compose; P byte-identical (7479a106); tests green.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(B): implicit weighted bi-prediction → adaptive B-count in `--…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-20T16:00:49Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "0a2b45029501c985f59084dc00dbd432daffb4da",
          "body": "…EFAULT-ON\n\nThe last lever. AQ's premise (busy = maskable) fails on pathological synthetic\ncontent (flat regions beside SALIENT detailed patterns), costing PSNR there. Fix: a\ncontent-adaptive EFFECTIVE strength that backs off as the frame's log-variance SPREAD\nrises — the same variance already compu\n[…]\ns the fifth knob driven by the\none per-GOP/per-frame content signal, and the first quality feature to ship on by\ndefault this arc.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(AQ): content-adaptive strength → AQ now regresses nowhere → D…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-20T15:34:16Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "9824f364b4eb6aeeec05e343fa1d879cee004d5c",
          "body": "AQ nets a rate change at fixed QP (coarsening a busy MB saves more bits than fining a\nflat one adds), so `--qp X` drifted off its expected rate and AQ couldn't drive rate\ncontrol. Fix: a per-frame base-QP shift `c = 6·log2(Σ var·2^(−dqp/6) / Σ var)` (a\nvariance-weighted `bits ∝ var·2^(−qp/6)` model)\n[…]\n materially stronger feature. Bit-exact vs ffmpeg (incl. with the B pipeline);\naq=0 still byte-identical (7479a106); still opt-in.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(AQ): rate-compensate the base QP — and it IMPROVES the BD-rate",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-20T15:24:19Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "6a21da09937d2736b273fbb62e983817efd41dea",
          "body": "Modulates the QP per macroblock: flat/low-variance MBs (where blocking & banding\nshow) get a FINER QP, busy/high-variance MBs (where the eye masks error) a COARSER\none — moving bits to where they're seen. The shift is `strength·(log2 var − frame\nmean log2 var)`, relative to THIS frame's texture dist\n[…]\nif it wins\neverywhere\" rule — it ships off, `--aq 1.0` recommended for natural video. Composes\nwith P/I and B-frames; tests green.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder: Adaptive Quantization (per-MB QP by content, opt-in)",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-20T02:27:51Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "dd020b0b7ab288353de06f120dc2f5ae6aa88fc3",
          "body": "…ive)\n\nThe fourth knob on the one per-GOP signal. B-frames are non-reference, so on a VERY\npredictable GOP (bi-pred + spatial-direct nail them → tiny residual) they can be\nquantized much harder for near-free bits.\n\nThe finding that shaped it: unlike the I-cascade's MILD content-dependence, the\nB-off\n[…]\n be dispatched aggressively). Busy B-GOPs untouched; P-only escape hatch\n(7479a106) intact; dpan bit-exact vs ffmpeg; tests green.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder: content-adaptive per-GOP B-frame QP offset (steep, conservat…",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-20T01:40:57Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "abb737fab802307437988a355eb7937ed1d493b3",
          "body": "Makes the ip_ratio cascade CONTENT-ADAPTIVE per GOP, driven by the same cheap signal\nas the B/P dispatch. On a predictable GOP the I-frame is a large FRACTION of the GOP's\nbits (P/B are cheap), so investing in it pays outsized — such GOPs deepen the I-offset\nup to 2 further QP steps; busy GOPs keep \n[…]\n+0.67%)\n  clip240 (busy)      unchanged (correctly selects −3, no bonus)\nConformant (dpan/mixed bit-exact vs ffmpeg); tests green.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder: content-adaptive per-GOP I-frame QP cascade",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-20T01:04:27Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "e39bede0a77ffc316ed311a79bb24d6f3744d4ce",
          "body": "The I-frame at each GOP boundary is the root reference for its whole GOP, so coding\nit FINER propagates quality to every P/B that predicts from it — the classic\n`ip_ratio`. Added as `i_qp_offset` (constant-QP), applied per-GOP in both the P\nstreaming path and the B-capable `code_picture`, so it comp\n[…]\n— the long-standing P sentinel, now the opt-out\nbaseline). All paths stay conformant (P/B/mixed bit-exact vs ffmpeg); tests green.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder: per-GOP I-frame QP cascade (ip_ratio), calibrated default -3",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-19T17:01:42Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "944ebb3d0584b2fc354f5b9dec09e17a4a498207",
          "body": "Refines `--bframes auto` from a whole-clip decision to PER-GOP, so a mixed clip\ngets B-frames on its smooth segments and P on its busy ones (the natural next grain\nof codec-content-adaptive-dispatch).\n\n- The favorability signal (`bframes_favorable`) now runs per GOP; a favorable GOP\n  uses the B str\n[…]\nsmooth clip → all-B (-19.6%), busy clip → P-only.\nConformant (mixed stream bit-exact vs ffmpeg); P path byte-identical (7479a106).\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(B): per-GOP content-adaptive B-frame dispatch",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-19T16:23:02Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "320a5239060f6cef5189bacd51cebdcf912aa2ba",
          "body": "B-frames are content-dependent — a big win on smooth/predictable motion (-19.6%\nBD-rate) but a loss on busy content (+3.6%). This dispatches per-CLIP so we capture\nthe win and never regress, following the content-adaptive-dispatch pattern:\n\n- Cheap O(pixels) content signal: `bframes_favorable()` — t\n[…]\n3.6% regression), each bit-exact-conformant vs ffmpeg. P path\nbyte-identical (sha 7479a106); `--bframes N` unchanged; default off.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(B): content-adaptive B-frame enable (`--bframes auto`)",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-19T15:57:31Z",
          "body_truncated": true,
          "is_coding_agent": true
        },
        {
          "oid": "67b1ddc1ef5b3dd315b993913d4ac4ae3ecaabd8",
          "body": "…B-mode set\n\nAdds the zero-mvd spatial-direct modes, gated bit-exact vs ffmpeg AND our decoder:\n\n- RefFrame now carries the per-4×4 List-0 motion field (mv/ref_idx/w4), populated\n  by anchors, so a later B-frame can read RefPicList1[0]'s co-located motion for\n  the colZeroFlag (mirrors the decoder's\n[…]\n. Feature is COMPLETE + CONFORMANT; not yet BD-rate-competitive.\nP path byte-identical (sha 7479a106); opt-in (bframes=0 default).\n\nCo-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>",
          "is_bot": false,
          "headline": "encoder(B): B_Direct_16x16 + B_Skip (spatial-direct) — completes the …",
          "author_name": "Tim",
          "author_login": "tim-almond-house",
          "committed_at": "2026-07-19T15:31:26Z",
          "body_truncated": true,
          "is_coding_agent": true
        }
      ],
      "releases_count": 0,
      "commits_last_year": 296,
      "latest_release_at": null,
      "latest_release_tag": null,
      "releases_from_tags": false,
      "days_since_last_push": 0,
      "active_weeks_last_year": 4,
      "days_since_latest_release": null,
      "mean_days_between_releases": null
    },
    "community": {
      "has_readme": true,
      "has_license": true,
      "has_description": true,
      "has_contributing": false,
      "health_percentage": 37,
      "has_issue_template": false,
      "has_code_of_conduct": false,
      "has_pull_request_template": false
    },
    "ecosystem": {
      "packages": [
        {
          "name": "rusty_h264",
          "exists": true,
          "license": "BSD-2-Clause",
          "keywords": [
            "avc",
            "codec",
            "decoder",
            "h264",
            "video",
            "multimedia::encoding",
            "multimedia::video"
          ],
          "ecosystem": "crates",
          "matches_repo": true,
          "registry_url": "https://crates.io/crates/rusty_h264",
          "is_deprecated": false,
          "latest_version": "0.4.0",
          "repository_url": "https://github.com/remade-with-rust/rusty_h264",
          "versions_count": 4,
          "total_downloads": 12783,
          "dependents_count": null,
          "deprecation_note": null,
          "maintainers_count": null,
          "monthly_downloads": 4261,
          "first_published_at": "2026-06-27T17:03:55.208375Z",
          "latest_published_at": "2026-07-30T05:00:56.657017Z",
          "latest_version_yanked": false,
          "days_since_latest_publish": 0
        },
        {
          "name": "rusty_h264-accel",
          "exists": true,
          "license": "BSD-2-Clause",
          "keywords": [
            "codec",
            "h264",
            "simd",
            "video",
            "hardware-support",
            "multimedia::video"
          ],
          "ecosystem": "crates",
          "matches_repo": true,
          "registry_url": "https://crates.io/crates/rusty_h264-accel",
          "is_deprecated": false,
          "latest_version": "0.4.0",
          "repository_url": "https://github.com/remade-with-rust/rusty_h264",
          "versions_count": 5,
          "total_downloads": 231,
          "dependents_count": null,
          "deprecation_note": null,
          "maintainers_count": null,
          "monthly_downloads": 77,
          "first_published_at": "2026-06-27T16:48:51.092825Z",
          "latest_published_at": "2026-07-30T05:00:13.174796Z",
          "latest_version_yanked": false,
          "days_since_latest_publish": 0
        },
        {
          "name": "rusty_h264-common",
          "exists": true,
          "license": "BSD-2-Clause",
          "keywords": [
            "avc",
            "bitstream",
            "codec",
            "h264",
            "video",
            "multimedia::video",
            "multimedia::encoding",
            "encoding"
          ],
          "ecosystem": "crates",
          "matches_repo": true,
          "registry_url": "https://crates.io/crates/rusty_h264-common",
          "is_deprecated": false,
          "latest_version": "0.4.0",
          "repository_url": "https://github.com/remade-with-rust/rusty_h264",
          "versions_count": 4,
          "total_downloads": 12927,
          "dependents_count": null,
          "deprecation_note": null,
          "maintainers_count": null,
          "monthly_downloads": 4309,
          "first_published_at": "2026-06-27T17:03:34.257981Z",
          "latest_published_at": "2026-07-30T05:00:20.702876Z",
          "latest_version_yanked": false,
          "days_since_latest_publish": 0
        },
        {
          "name": "rusty_h264-decoder",
          "exists": true,
          "license": "BSD-2-Clause",
          "keywords": [
            "avc",
            "codec",
            "decoder",
            "h264",
            "video",
            "multimedia::encoding",
            "multimedia::video"
          ],
          "ecosystem": "crates",
          "matches_repo": true,
          "registry_url": "https://crates.io/crates/rusty_h264-decoder",
          "is_deprecated": false,
          "latest_version": "0.4.0",
          "repository_url": "https://github.com/remade-with-rust/rusty_h264",
          "versions_count": 4,
          "total_downloads": 12798,
          "dependents_count": null,
          "deprecation_note": null,
          "maintainers_count": null,
          "monthly_downloads": 4266,
          "first_published_at": "2026-06-27T17:03:47.822335Z",
          "latest_published_at": "2026-07-30T05:00:40.251060Z",
          "latest_version_yanked": false,
          "days_since_latest_publish": 0
        },
        {
          "name": "rusty_h264-encoder",
          "exists": true,
          "license": "BSD-2-Clause",
          "keywords": [
            "avc",
            "codec",
            "encoder",
            "h264",
            "video",
            "multimedia::encoding",
            "multimedia::video"
          ],
          "ecosystem": "crates",
          "matches_repo": true,
          "registry_url": "https://crates.io/crates/rusty_h264-encoder",
          "is_deprecated": false,
          "latest_version": "0.4.0",
          "repository_url": "https://github.com/remade-with-rust/rusty_h264",
          "versions_count": 4,
          "total_downloads": 12903,
          "dependents_count": null,
          "deprecation_note": null,
          "maintainers_count": null,
          "monthly_downloads": 4301,
          "first_published_at": "2026-06-27T17:03:41.534818Z",
          "latest_published_at": "2026-07-30T05:00:31.602964Z",
          "latest_version_yanked": false,
          "days_since_latest_publish": 0
        }
      ]
    },
    "popularity": {
      "forks": 0,
      "stars": 0,
      "watchers": 0,
      "fork_history": {
        "days": [],
        "complete": true,
        "collected": 0,
        "total_forks": 0
      },
      "star_history": {
        "days": [],
        "complete": true,
        "collected": 0,
        "total_stars": 0,
        "collected_at": null
      },
      "open_issues_and_prs": 0
    },
    "ai_readiness": {
      "has_nix": false,
      "example_dirs": [
        "examples"
      ],
      "has_llms_txt": false,
      "has_dockerfile": false,
      "has_mcp_signal": false,
      "bootstrap_files": [],
      "api_schema_files": [],
      "has_devcontainer": false,
      "typecheck_configs": [],
      "toolchain_manifests": [
        "Cargo.toml",
        "bench/Cargo.toml",
        "crates/rusty_h264-accel/Cargo.toml",
        "crates/rusty_h264-cli/Cargo.toml",
        "crates/rusty_h264-common/Cargo.toml",
        "crates/rusty_h264-decoder/Cargo.toml",
        "crates/rusty_h264-encoder/Cargo.toml",
        "crates/rusty_h264/Cargo.toml",
        "video-tests/analyzer/Cargo.toml"
      ],
      "largest_source_bytes": 375018,
      "source_files_sampled": 85,
      "oversized_source_files": 4,
      "agent_instruction_files": [],
      "agent_instruction_max_bytes": null
    },
    "dependencies": {
      "manifests": [
        "Cargo.toml",
        "bench/Cargo.toml"
      ],
      "advisories": {
        "error": null,
        "scope": null,
        "source": null,
        "findings": [],
        "collected": false,
        "malicious": [],
        "truncated": false,
        "by_severity": {},
        "advisory_count": 0,
        "affected_count": 0,
        "assessed_count": 0,
        "malicious_count": 0,
        "assessed_package": null,
        "unassessed_count": 0,
        "direct_affected_count": 0
      },
      "ecosystems": [
        "crates"
      ],
      "dependencies": [
        {
          "name": "rusty_h264",
          "manifest": "bench/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": null
        },
        {
          "name": "rusty_h264",
          "manifest": "crates/rusty_h264-cli/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": "0.4.0"
        },
        {
          "name": "rusty_h264-common",
          "manifest": "crates/rusty_h264-cli/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": null
        },
        {
          "name": "wide",
          "manifest": "crates/rusty_h264-common/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": null
        },
        {
          "name": "bytemuck",
          "manifest": "crates/rusty_h264-common/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": "1"
        },
        {
          "name": "rusty_h264-accel",
          "manifest": "crates/rusty_h264-common/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": "0.4.0"
        },
        {
          "name": "rusty_h264-common",
          "manifest": "crates/rusty_h264-decoder/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": null
        },
        {
          "name": "rusty_h264-accel",
          "manifest": "crates/rusty_h264-decoder/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": "0.4.0"
        },
        {
          "name": "rusty_h264-common",
          "manifest": "crates/rusty_h264-encoder/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": null
        },
        {
          "name": "rusty_h264-accel",
          "manifest": "crates/rusty_h264-encoder/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": "0.4.0"
        },
        {
          "name": "rusty_h264-common",
          "manifest": "crates/rusty_h264/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": null
        },
        {
          "name": "rusty_h264-encoder",
          "manifest": "crates/rusty_h264/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": null
        },
        {
          "name": "rusty_h264-decoder",
          "manifest": "crates/rusty_h264/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": null
        },
        {
          "name": "rusty_h264-common",
          "manifest": "video-tests/analyzer/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": null
        },
        {
          "name": "rusty_h264-encoder",
          "manifest": "video-tests/analyzer/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": null
        },
        {
          "name": "rusty_h264-decoder",
          "manifest": "video-tests/analyzer/Cargo.toml",
          "ecosystem": "crates",
          "version_constraint": null
        }
      ],
      "all_dependencies": {
        "error": "GitHub dependency-graph SBOM unavailable (404); the dependency graph may be disabled for this repository",
        "source": null,
        "packages": [],
        "collected": false,
        "truncated": false,
        "total_count": null,
        "direct_count": null,
        "indirect_count": null
      }
    },
    "maintainership": {
      "issues": {
        "open_prs": 0,
        "merged_prs": 0,
        "open_issues": 0,
        "closed_ratio": null,
        "closed_issues": 0,
        "closed_unmerged_prs": 0
      },
      "bus_factor": 1,
      "bot_contributors": 0,
      "top_contributors": [
        {
          "type": "User",
          "login": "tim-almond-house",
          "commits": 317,
          "avatar_url": "https://avatars.githubusercontent.com/u/274574578?v=4"
        }
      ],
      "contributors_sampled": 1,
      "top_contributor_share": 1
    },
    "quality_signals": {
      "has_ci": true,
      "has_tests": true,
      "ci_workflows": [
        "ci.yml"
      ],
      "has_docs_dir": true,
      "linter_configs": [],
      "has_editorconfig": false,
      "has_linter_config": false,
      "has_precommit_config": false
    },
    "security_signals": {
      "lockfiles": [],
      "scorecard": {
        "checks": [
          {
            "name": "Binary-Artifacts",
            "score": 10,
            "reason": "no binaries found in the repo",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#binary-artifacts"
          },
          {
            "name": "Branch-Protection",
            "score": 0,
            "reason": "branch protection not enabled on development/release branches",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#branch-protection"
          },
          {
            "name": "CI-Tests",
            "score": null,
            "reason": "no pull request found",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#ci-tests"
          },
          {
            "name": "CII-Best-Practices",
            "score": 0,
            "reason": "no effort to earn an OpenSSF best practices badge detected",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#cii-best-practices"
          },
          {
            "name": "Code-Review",
            "score": 0,
            "reason": "Found 0/30 approved changesets -- score normalized to 0",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#code-review"
          },
          {
            "name": "Contributors",
            "score": 0,
            "reason": "project has 0 contributing companies or organizations -- score normalized to 0",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#contributors"
          },
          {
            "name": "Dangerous-Workflow",
            "score": 10,
            "reason": "no dangerous workflow patterns detected",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#dangerous-workflow"
          },
          {
            "name": "Dependency-Update-Tool",
            "score": 0,
            "reason": "no update tool detected",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#dependency-update-tool"
          },
          {
            "name": "Fuzzing",
            "score": 0,
            "reason": "project is not fuzzed",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#fuzzing"
          },
          {
            "name": "License",
            "score": 10,
            "reason": "license file detected",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#license"
          },
          {
            "name": "Maintained",
            "score": 0,
            "reason": "project was created within the last 90 days. Please review its contents carefully",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#maintained"
          },
          {
            "name": "Packaging",
            "score": null,
            "reason": "packaging workflow not detected",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#packaging"
          },
          {
            "name": "Pinned-Dependencies",
            "score": 0,
            "reason": "dependency not pinned by hash detected -- score normalized to 0",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#pinned-dependencies"
          },
          {
            "name": "SAST",
            "score": 0,
            "reason": "no SAST tool detected",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#sast"
          },
          {
            "name": "Security-Policy",
            "score": 0,
            "reason": "security policy file not detected",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#security-policy"
          },
          {
            "name": "Signed-Releases",
            "score": null,
            "reason": "no releases found",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#signed-releases"
          },
          {
            "name": "Token-Permissions",
            "score": 0,
            "reason": "detected GitHub workflow tokens with excessive permissions",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#token-permissions"
          },
          {
            "name": "Vulnerabilities",
            "score": 10,
            "reason": "0 existing vulnerabilities detected",
            "documentation_url": "https://github.com/ossf/scorecard/blob/c395761df6afe1a69e476bc60a013a94bcbc153f/docs/checks.md#vulnerabilities"
          }
        ],
        "commit": "b512bcb88f40dfad298251e274643fb95bcb359a",
        "ran_at": "2026-07-30T10:39:18Z",
        "aggregate_score": 3.1,
        "scorecard_version": "v5.5.0"
      },
      "has_codeql_workflow": false,
      "has_security_policy": false,
      "has_dependabot_config": false
    },
    "contribution_flow": {
      "collected": true,
      "ci_last_run_at": "2026-07-30T04:59:07Z",
      "oldest_open_prs": [],
      "last_merged_pr_at": null,
      "ci_last_conclusion": "FAILURE",
      "oldest_open_issues": []
    }
  },
  "config": {
    "disabled_metrics": [],
    "disabled_categories": [],
    "disabled_components": {}
  },
  "source": {
    "url": "https://github.com/Remade-With-Rust/rusty_h264",
    "host": "github.com",
    "name": "rusty_h264",
    "owner": "Remade-With-Rust"
  },
  "metrics": {
    "overall": {
      "key": "overall",
      "band": "at_risk",
      "name": "Overall health",
      "note": null,
      "notes": [],
      "value": 41,
      "inputs": {
        "security": 31,
        "vitality": 35,
        "community": 36,
        "governance": 33,
        "engineering": 72
      },
      "components": []
    },
    "categories": [
      {
        "key": "vitality",
        "band": "at_risk",
        "name": "Vitality",
        "value": 35,
        "weight": 0.22,
        "metrics": [
          {
            "key": "development_activity",
            "band": "moderate",
            "name": "Development activity",
            "note": null,
            "notes": [],
            "value": 57,
            "inputs": {
              "commits_last_year": 296,
              "human_commit_share": 1,
              "days_since_last_push": 0,
              "active_weeks_last_year": 4
            },
            "components": [
              {
                "key": "push_recency",
                "name": "Push recency",
                "detail": "last push 0 days ago",
                "points": 36,
                "status": "met",
                "details": [
                  {
                    "code": "push_recency",
                    "params": {
                      "days": 0
                    }
                  }
                ],
                "max_points": 36
              },
              {
                "key": "commit_cadence",
                "name": "Commit cadence",
                "detail": "4/52 weeks with commits",
                "points": 2.8,
                "status": "partial",
                "details": [
                  {
                    "code": "commit_cadence_weeks",
                    "params": {
                      "weeks": 4
                    }
                  }
                ],
                "max_points": 36
              },
              {
                "key": "commit_volume",
                "name": "Commit volume",
                "detail": "296 commits in the last year",
                "points": 18,
                "status": "met",
                "details": [
                  {
                    "code": "commits_last_year",
                    "params": {
                      "count": 296
                    }
                  }
                ],
                "max_points": 18
              },
              {
                "key": "openssf_scorecard_maintained",
                "name": "OpenSSF Scorecard: Maintained",
                "detail": "project was created within the last 90 days. Please review its contents carefully",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 10
              }
            ]
          },
          {
            "key": "release_discipline",
            "band": "critical",
            "name": "Release discipline",
            "note": "Excluded from scoring (no data or not applicable): OpenSSF Scorecard: Signed-Releases. Remaining weights renormalized.",
            "notes": [
              {
                "code": "excluded_no_data",
                "params": {
                  "components": [
                    "openssf_scorecard_signed_releases"
                  ]
                }
              },
              {
                "code": "weights_renormalized",
                "params": {}
              }
            ],
            "value": 1,
            "inputs": {
              "releases_count": 0
            },
            "components": [
              {
                "key": "ships_releases",
                "name": "Ships releases",
                "detail": "no releases published",
                "points": 0,
                "status": "missed",
                "details": [
                  {
                    "code": "no_releases_published",
                    "params": {}
                  }
                ],
                "max_points": 27
              },
              {
                "key": "release_recency",
                "name": "Release recency",
                "detail": "no releases",
                "points": 0,
                "status": "missed",
                "details": [
                  {
                    "code": "no_releases",
                    "params": {}
                  }
                ],
                "max_points": 36
              },
              {
                "key": "release_cadence",
                "name": "Release cadence",
                "detail": "no releases",
                "points": 0,
                "status": "missed",
                "details": [
                  {
                    "code": "no_releases",
                    "params": {}
                  }
                ],
                "max_points": 27
              },
              {
                "key": "openssf_scorecard_signed_releases",
                "name": "OpenSSF Scorecard: Signed-Releases",
                "detail": "no releases found",
                "points": 0,
                "status": "excluded",
                "details": [
                  {
                    "code": "no_data",
                    "params": {}
                  }
                ],
                "max_points": 10
              }
            ]
          },
          {
            "key": "abandonment",
            "band": "excellent",
            "name": "Abandonment",
            "note": null,
            "notes": [],
            "value": 100,
            "inputs": {
              "cap": null,
              "state": "unverified",
              "guards": [],
              "signals": [],
              "red_flag": false,
              "multiplier_pct": 100,
              "declared_reason": null,
              "unverified_reason": "repository_too_young",
              "unanswered_open_prs": null,
              "unanswered_open_issues": null,
              "days_since_last_merged_pr": null,
              "days_since_last_human_commit": null,
              "days_since_last_human_commit_is_floor": false
            },
            "components": [
              {
                "key": "project_is_still_maintained",
                "name": "Project is still maintained",
                "detail": "maintenance record not established from the collected data",
                "points": 100,
                "status": "met",
                "details": [
                  {
                    "code": "abandonment_unverified",
                    "params": {}
                  }
                ],
                "max_points": 100
              }
            ]
          }
        ],
        "description": "Is the project alive — is code being written and are releases shipping?"
      },
      {
        "key": "community",
        "band": "at_risk",
        "name": "Community & Adoption",
        "value": 36,
        "weight": 0.18,
        "metrics": [
          {
            "key": "popularity",
            "band": "critical",
            "name": "Popularity & adoption",
            "note": null,
            "notes": [],
            "value": 1,
            "inputs": {
              "forks": 0,
              "stars": 0,
              "watchers": 0,
              "growth_state": "unverified",
              "growth_factor_pct": 100,
              "growth_unverified_reason": "no_history"
            },
            "components": [
              {
                "key": "stars",
                "name": "Stars",
                "detail": "0 stars",
                "points": 0,
                "status": "missed",
                "details": [
                  {
                    "code": "stars",
                    "params": {
                      "count": 0
                    }
                  }
                ],
                "max_points": 60
              },
              {
                "key": "forks",
                "name": "Forks",
                "detail": "0 forks",
                "points": 0,
                "status": "missed",
                "details": [
                  {
                    "code": "forks",
                    "params": {
                      "count": 0
                    }
                  }
                ],
                "max_points": 25
              },
              {
                "key": "watchers",
                "name": "Watchers",
                "detail": "0 watchers",
                "points": 0,
                "status": "missed",
                "details": [
                  {
                    "code": "watchers",
                    "params": {
                      "count": 0
                    }
                  }
                ],
                "max_points": 15
              }
            ]
          },
          {
            "key": "community_health",
            "band": "moderate",
            "name": "Community health",
            "note": null,
            "notes": [],
            "value": 50,
            "inputs": {
              "has_readme": true,
              "has_license": true,
              "has_contributing": false,
              "has_issue_template": false,
              "has_code_of_conduct": false,
              "has_pull_request_template": false
            },
            "components": [
              {
                "key": "readme",
                "name": "README",
                "detail": null,
                "points": 22.5,
                "status": "met",
                "details": [],
                "max_points": 22.5
              },
              {
                "key": "license",
                "name": "License",
                "detail": "recognized license (BSD-2-Clause)",
                "points": 22.5,
                "status": "met",
                "details": [
                  {
                    "code": "license_standard",
                    "params": {}
                  },
                  {
                    "code": "license_spdx",
                    "params": {
                      "spdx": "BSD-2-Clause"
                    }
                  }
                ],
                "max_points": 22.5
              },
              {
                "key": "contributing_guide",
                "name": "CONTRIBUTING guide",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 18
              },
              {
                "key": "code_of_conduct",
                "name": "Code of conduct",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 13.5
              },
              {
                "key": "issue_template",
                "name": "Issue template",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 7.2
              },
              {
                "key": "pr_template",
                "name": "PR template",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 6.3
              }
            ]
          },
          {
            "key": "ecosystem_adoption",
            "band": "good",
            "name": "Ecosystem adoption (downloads)",
            "note": "Excluded from scoring (no data or not applicable): Registry dependents. Remaining weights renormalized.",
            "notes": [
              {
                "code": "excluded_no_data",
                "params": {
                  "components": [
                    "registry_dependents"
                  ]
                }
              },
              {
                "code": "weights_renormalized",
                "params": {}
              }
            ],
            "value": 71,
            "inputs": {
              "packages": [
                "rusty_h264",
                "rusty_h264-accel",
                "rusty_h264-common",
                "rusty_h264-decoder",
                "rusty_h264-encoder"
              ],
              "dependents": null,
              "ecosystems": "crates",
              "total_downloads": 51642,
              "monthly_downloads": 17214
            },
            "components": [
              {
                "key": "monthly_downloads",
                "name": "Monthly downloads",
                "detail": "17,214 downloads/month across crates",
                "points": 56.5,
                "status": "partial",
                "details": [
                  {
                    "code": "downloads_monthly",
                    "params": {
                      "count": 17214,
                      "ecosystems": "crates"
                    }
                  }
                ],
                "max_points": 80
              },
              {
                "key": "registry_dependents",
                "name": "Registry dependents",
                "detail": "not reported by this ecosystem",
                "points": 0,
                "status": "excluded",
                "details": [
                  {
                    "code": "not_reported_by_this_ecosystem",
                    "params": {}
                  }
                ],
                "max_points": 20
              }
            ]
          }
        ],
        "description": "Does the project have users, downloads, attention, and a welcoming setup for contributors?"
      },
      {
        "key": "governance",
        "band": "at_risk",
        "name": "Sustainability & Governance",
        "value": 33,
        "weight": 0.24,
        "metrics": [
          {
            "key": "maintainer_resilience",
            "band": "critical",
            "name": "Maintainer resilience (bus factor)",
            "note": null,
            "notes": [],
            "value": 10,
            "inputs": {
              "bus_factor": 1,
              "contributors_sampled": 1,
              "top_contributor_share": 1
            },
            "components": [
              {
                "key": "bus_factor",
                "name": "Bus factor",
                "detail": "1 contributor(s) cover half of all commits",
                "points": 9,
                "status": "partial",
                "details": [
                  {
                    "code": "bus_factor",
                    "params": {
                      "count": 1
                    }
                  }
                ],
                "max_points": 54
              },
              {
                "key": "commit_distribution",
                "name": "Commit distribution",
                "detail": "top contributor authored 100% of commits",
                "points": 0,
                "status": "missed",
                "details": [
                  {
                    "code": "top_contributor_share",
                    "params": {
                      "share": 100
                    }
                  }
                ],
                "max_points": 22.5
              },
              {
                "key": "contributor_breadth",
                "name": "Contributor breadth",
                "detail": "1 contributors",
                "points": 1.4,
                "status": "partial",
                "details": [
                  {
                    "code": "contributors_sampled",
                    "params": {
                      "count": 1
                    }
                  }
                ],
                "max_points": 13.5
              },
              {
                "key": "openssf_scorecard_contributors",
                "name": "OpenSSF Scorecard: Contributors",
                "detail": "project has 0 contributing companies or organizations -- score normalized to 0",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 10
              }
            ]
          },
          {
            "key": "responsiveness",
            "band": "critical",
            "name": "Issue & PR responsiveness",
            "note": "Excluded from scoring (no data or not applicable): Issue resolution, PR acceptance. Remaining weights renormalized.",
            "notes": [
              {
                "code": "excluded_no_data",
                "params": {
                  "components": [
                    "issue_resolution",
                    "pr_acceptance"
                  ]
                }
              },
              {
                "code": "weights_renormalized",
                "params": {}
              }
            ],
            "value": 1,
            "inputs": {
              "merged_prs": 0,
              "open_issues": 0,
              "closed_issues": 0,
              "issue_closed_ratio": null,
              "closed_unmerged_prs": 0
            },
            "components": [
              {
                "key": "issue_resolution",
                "name": "Issue resolution",
                "detail": "no issues or no data",
                "points": 0,
                "status": "excluded",
                "details": [
                  {
                    "code": "no_issues_or_data",
                    "params": {}
                  }
                ],
                "max_points": 46.75
              },
              {
                "key": "pr_acceptance",
                "name": "PR acceptance",
                "detail": "no decided pull requests or no data",
                "points": 0,
                "status": "excluded",
                "details": [
                  {
                    "code": "no_decided_prs_or_data",
                    "params": {}
                  }
                ],
                "max_points": 38.25
              },
              {
                "key": "openssf_scorecard_code_review",
                "name": "OpenSSF Scorecard: Code-Review",
                "detail": "Found 0/30 approved changesets -- score normalized to 0",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 15
              }
            ]
          },
          {
            "key": "stewardship",
            "band": "at_risk",
            "name": "Ownership & stewardship",
            "note": null,
            "notes": [],
            "value": 40,
            "inputs": {
              "followers": 1,
              "owner_type": "Organization",
              "is_verified": null,
              "owner_login": "Remade-With-Rust",
              "public_repos": 11,
              "account_age_days": 39
            },
            "components": [
              {
                "key": "ownership_backing",
                "name": "Ownership backing",
                "detail": "organization-owned",
                "points": 30,
                "status": "met",
                "details": [
                  {
                    "code": "owner_organization",
                    "params": {}
                  }
                ],
                "max_points": 30
              },
              {
                "key": "verified_domain",
                "name": "Verified domain",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 20
              },
              {
                "key": "owner_reach",
                "name": "Owner reach",
                "detail": "1 followers of Remade-With-Rust",
                "points": 2.2,
                "status": "partial",
                "details": [
                  {
                    "code": "owner_followers",
                    "params": {
                      "count": 1,
                      "login": "Remade-With-Rust"
                    }
                  }
                ],
                "max_points": 25
              },
              {
                "key": "track_record",
                "name": "Track record",
                "detail": "11 public repos, account ~0 yr old",
                "points": 8.1,
                "status": "partial",
                "details": [
                  {
                    "code": "public_repos",
                    "params": {
                      "count": 11
                    }
                  },
                  {
                    "code": "account_age_years",
                    "params": {
                      "years": 0
                    }
                  }
                ],
                "max_points": 25
              }
            ]
          },
          {
            "key": "package_maintenance",
            "band": "excellent",
            "name": "Package maintenance",
            "note": null,
            "notes": [],
            "value": 100,
            "inputs": {
              "packages": [
                "rusty_h264",
                "rusty_h264-accel",
                "rusty_h264-common",
                "rusty_h264-decoder",
                "rusty_h264-encoder"
              ],
              "ecosystems": "crates",
              "any_deprecated": false,
              "min_days_since_publish": 0
            },
            "components": [
              {
                "key": "published_resolvable",
                "name": "Published & resolvable",
                "detail": "5 package(s) on crates",
                "points": 25,
                "status": "met",
                "details": [
                  {
                    "code": "packages_published",
                    "params": {
                      "count": 5,
                      "ecosystems": "crates"
                    }
                  }
                ],
                "max_points": 25
              },
              {
                "key": "publish_recency",
                "name": "Publish recency",
                "detail": "latest publish 0 days ago",
                "points": 35,
                "status": "met",
                "details": [
                  {
                    "code": "publish_recency",
                    "params": {
                      "days": 0
                    }
                  }
                ],
                "max_points": 35
              },
              {
                "key": "version_history",
                "name": "Version history",
                "detail": "5 published versions",
                "points": 20,
                "status": "met",
                "details": [
                  {
                    "code": "published_versions",
                    "params": {
                      "count": 5
                    }
                  }
                ],
                "max_points": 20
              },
              {
                "key": "not_deprecated",
                "name": "Not deprecated",
                "detail": "active, not deprecated or yanked",
                "points": 20,
                "status": "met",
                "details": [
                  {
                    "code": "package_not_deprecated",
                    "params": {}
                  }
                ],
                "max_points": 20
              }
            ]
          }
        ],
        "description": "Will the project survive its people — bus factor, responsiveness, who backs it, and package upkeep?"
      },
      {
        "key": "engineering",
        "band": "good",
        "name": "Engineering Quality",
        "value": 72,
        "weight": 0.2,
        "metrics": [
          {
            "key": "engineering_practices",
            "band": "moderate",
            "name": "Engineering practices",
            "note": "Excluded from scoring (no data or not applicable): OpenSSF Scorecard: CI-Tests. Remaining weights renormalized.",
            "notes": [
              {
                "code": "excluded_no_data",
                "params": {
                  "components": [
                    "openssf_scorecard_ci_tests"
                  ]
                }
              },
              {
                "code": "weights_renormalized",
                "params": {}
              }
            ],
            "value": 60,
            "inputs": {
              "has_ci": true,
              "has_tests": true,
              "has_editorconfig": false,
              "has_linter_config": false,
              "has_precommit_config": false
            },
            "components": [
              {
                "key": "ci_workflows",
                "name": "CI workflows",
                "detail": "1 workflow(s)",
                "points": 24,
                "status": "met",
                "details": [
                  {
                    "code": "ci_workflows",
                    "params": {
                      "count": 1
                    }
                  }
                ],
                "max_points": 24
              },
              {
                "key": "tests_present",
                "name": "Tests present",
                "detail": null,
                "points": 24,
                "status": "met",
                "details": [],
                "max_points": 24
              },
              {
                "key": "linter_config",
                "name": "Linter config",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 16
              },
              {
                "key": "pre_commit_hooks",
                "name": "Pre-commit hooks",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 9.6
              },
              {
                "key": "editorconfig",
                "name": ".editorconfig",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 6.4
              },
              {
                "key": "openssf_scorecard_ci_tests",
                "name": "OpenSSF Scorecard: CI-Tests",
                "detail": "no pull request found",
                "points": 0,
                "status": "excluded",
                "details": [
                  {
                    "code": "no_data",
                    "params": {}
                  }
                ],
                "max_points": 20
              }
            ]
          },
          {
            "key": "documentation",
            "band": "excellent",
            "name": "Documentation",
            "note": null,
            "notes": [],
            "value": 90,
            "inputs": {
              "topics": [
                "built-with-rust",
                "codec",
                "ffmpeg",
                "rust",
                "embedded"
              ],
              "has_wiki": false,
              "homepage": "https://www.mata.network/remade-with-rust",
              "has_readme": true,
              "has_docs_dir": true,
              "has_description": true
            },
            "components": [
              {
                "key": "readme",
                "name": "README",
                "detail": null,
                "points": 30,
                "status": "met",
                "details": [],
                "max_points": 30
              },
              {
                "key": "documentation_directory",
                "name": "Documentation directory",
                "detail": null,
                "points": 25,
                "status": "met",
                "details": [],
                "max_points": 25
              },
              {
                "key": "documentation_homepage_site",
                "name": "Documentation / homepage site",
                "detail": "https://www.mata.network/remade-with-rust",
                "points": 15,
                "status": "met",
                "details": [],
                "max_points": 15
              },
              {
                "key": "repository_description",
                "name": "Repository description",
                "detail": null,
                "points": 10,
                "status": "met",
                "details": [],
                "max_points": 10
              },
              {
                "key": "topics",
                "name": "Topics",
                "detail": "5 topics",
                "points": 10,
                "status": "met",
                "details": [
                  {
                    "code": "topics_count",
                    "params": {
                      "count": 5
                    }
                  }
                ],
                "max_points": 10
              },
              {
                "key": "wiki",
                "name": "Wiki",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 10
              }
            ]
          }
        ],
        "description": "Are baseline engineering and documentation practices in place?"
      },
      {
        "key": "security",
        "band": "at_risk",
        "name": "Security",
        "value": 31,
        "weight": 0.16,
        "metrics": [
          {
            "key": "security_posture",
            "band": "at_risk",
            "name": "Security posture",
            "note": "Excluded from scoring (no data or not applicable): CI-Tests, Packaging, Signed-Releases. Remaining weights renormalized.",
            "notes": [
              {
                "code": "excluded_no_data",
                "params": {
                  "components": [
                    "ci_tests",
                    "packaging",
                    "signed_releases"
                  ]
                }
              },
              {
                "code": "weights_renormalized",
                "params": {}
              }
            ],
            "value": 31,
            "inputs": {
              "source": "openssf_scorecard",
              "checks_evaluated": 15,
              "scorecard_version": "v5.5.0",
              "checks_inconclusive": 3,
              "scorecard_aggregate": 3.1
            },
            "components": [
              {
                "key": "binary_artifacts",
                "name": "Binary-Artifacts",
                "detail": "no binaries found in the repo",
                "points": 7.5,
                "status": "met",
                "details": [],
                "max_points": 7.5
              },
              {
                "key": "branch_protection",
                "name": "Branch-Protection",
                "detail": "branch protection not enabled on development/release branches",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 7.5
              },
              {
                "key": "ci_tests",
                "name": "CI-Tests",
                "detail": "no pull request found",
                "points": 0,
                "status": "excluded",
                "details": [
                  {
                    "code": "no_data",
                    "params": {}
                  }
                ],
                "max_points": 2.5
              },
              {
                "key": "cii_best_practices",
                "name": "CII-Best-Practices",
                "detail": "no effort to earn an OpenSSF best practices badge detected",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 2.5
              },
              {
                "key": "code_review",
                "name": "Code-Review",
                "detail": "Found 0/30 approved changesets -- score normalized to 0",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 7.5
              },
              {
                "key": "contributors",
                "name": "Contributors",
                "detail": "project has 0 contributing companies or organizations -- score normalized to 0",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 2.5
              },
              {
                "key": "dangerous_workflow",
                "name": "Dangerous-Workflow",
                "detail": "no dangerous workflow patterns detected",
                "points": 10,
                "status": "met",
                "details": [],
                "max_points": 10
              },
              {
                "key": "dependency_update_tool",
                "name": "Dependency-Update-Tool",
                "detail": "no update tool detected",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 7.5
              },
              {
                "key": "fuzzing",
                "name": "Fuzzing",
                "detail": "project is not fuzzed",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 5
              },
              {
                "key": "license",
                "name": "License",
                "detail": "license file detected",
                "points": 2.5,
                "status": "met",
                "details": [],
                "max_points": 2.5
              },
              {
                "key": "maintained",
                "name": "Maintained",
                "detail": "project was created within the last 90 days. Please review its contents carefully",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 7.5
              },
              {
                "key": "packaging",
                "name": "Packaging",
                "detail": "packaging workflow not detected",
                "points": 0,
                "status": "excluded",
                "details": [
                  {
                    "code": "no_data",
                    "params": {}
                  }
                ],
                "max_points": 5
              },
              {
                "key": "pinned_dependencies",
                "name": "Pinned-Dependencies",
                "detail": "dependency not pinned by hash detected -- score normalized to 0",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 5
              },
              {
                "key": "sast",
                "name": "SAST",
                "detail": "no SAST tool detected",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 5
              },
              {
                "key": "security_policy",
                "name": "Security-Policy",
                "detail": "security policy file not detected",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 5
              },
              {
                "key": "signed_releases",
                "name": "Signed-Releases",
                "detail": "no releases found",
                "points": 0,
                "status": "excluded",
                "details": [
                  {
                    "code": "no_data",
                    "params": {}
                  }
                ],
                "max_points": 7.5
              },
              {
                "key": "token_permissions",
                "name": "Token-Permissions",
                "detail": "detected GitHub workflow tokens with excessive permissions",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 7.5
              },
              {
                "key": "vulnerabilities",
                "name": "Vulnerabilities",
                "detail": "0 existing vulnerabilities detected",
                "points": 7.5,
                "status": "met",
                "details": [],
                "max_points": 7.5
              }
            ]
          },
          {
            "key": "high_risk_jurisdiction_exposure",
            "band": "excellent",
            "name": "High-Risk Jurisdiction Exposure",
            "note": "Only high-confidence self-published location evidence affects this multiplier. Ambiguous matches are review-only; country evidence is not proof of nationality, citizenship, legal registration, malicious intent, or sanctions status.",
            "notes": [
              {
                "code": "jurisdiction_evidence_limits",
                "params": {}
              }
            ],
            "value": 100,
            "inputs": {
              "meaning": "self-published location evidence; not nationality or citizenship",
              "red_flag": false,
              "exposures": [],
              "policy_countries": [
                "Russia",
                "Iran",
                "North Korea"
              ],
              "review_only_matches": 0,
              "assessed_self_published_locations": 1
            },
            "components": [
              {
                "key": "policy_exposure_multiplier",
                "name": "Policy exposure multiplier",
                "detail": "no confirmed policy-scope location match",
                "points": 100,
                "status": "met",
                "details": [
                  {
                    "code": "jurisdiction_no_match",
                    "params": {}
                  }
                ],
                "max_points": 100
              }
            ]
          }
        ],
        "description": "Are visible security and supply-chain practices strong, with no malicious dependency and no unresolved high-risk jurisdiction exposure?"
      },
      {
        "key": "ai_readiness",
        "band": "moderate",
        "name": "AI Readiness",
        "value": 55,
        "weight": 0,
        "metrics": [
          {
            "key": "ai_agent_context",
            "band": "at_risk",
            "name": "Agent context & guidance",
            "note": null,
            "notes": [],
            "value": 40,
            "inputs": {
              "has_llms_txt": false,
              "legible_history_share": 1,
              "agent_instruction_files": [],
              "agent_instruction_max_bytes": null
            },
            "components": [
              {
                "key": "agent_instructions",
                "name": "Agent instructions",
                "detail": "no CLAUDE.md / AGENTS.md / editor rules",
                "points": 0,
                "status": "missed",
                "details": [
                  {
                    "code": "no_agent_instructions",
                    "params": {}
                  }
                ],
                "max_points": 45
              },
              {
                "key": "machine_readable_docs_llms_txt",
                "name": "Machine-readable docs (llms.txt)",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 15
              },
              {
                "key": "legible_commit_history",
                "name": "Legible commit history",
                "detail": "100 of 100 human commits state their intent (structured subject or explanatory body)",
                "points": 40,
                "status": "met",
                "details": [
                  {
                    "code": "legible_history",
                    "params": {
                      "legible": 100,
                      "sampled": 100
                    }
                  }
                ],
                "max_points": 40
              }
            ]
          },
          {
            "key": "ai_verify_loop",
            "band": "moderate",
            "name": "Verify loop (build / test / typecheck)",
            "note": null,
            "notes": [],
            "value": 56,
            "inputs": {
              "has_nix": false,
              "has_tests": true,
              "lockfiles": [],
              "has_dockerfile": false,
              "typed_language": true,
              "bootstrap_files": [],
              "has_devcontainer": false,
              "has_linter_config": false,
              "typecheck_configs": [],
              "agent_commit_share": 1,
              "toolchain_manifests": [
                "Cargo.toml",
                "bench/Cargo.toml",
                "crates/rusty_h264-accel/Cargo.toml",
                "crates/rusty_h264-cli/Cargo.toml",
                "crates/rusty_h264-common/Cargo.toml",
                "crates/rusty_h264-decoder/Cargo.toml",
                "crates/rusty_h264-encoder/Cargo.toml",
                "crates/rusty_h264/Cargo.toml",
                "video-tests/analyzer/Cargo.toml"
              ],
              "dependency_bot_commit_share": 0
            },
            "components": [
              {
                "key": "one_command_bootstrap",
                "name": "One-command bootstrap",
                "detail": "Cargo.toml, bench/Cargo.toml, crates/rusty_h264-accel/Cargo.toml (toolchain convention, no task runner)",
                "points": 12.6,
                "status": "partial",
                "details": [
                  {
                    "code": "toolchain_convention",
                    "params": {
                      "files": "Cargo.toml, bench/Cargo.toml, crates/rusty_h264-accel/Cargo.toml"
                    }
                  }
                ],
                "max_points": 18
              },
              {
                "key": "automated_tests",
                "name": "Automated tests",
                "detail": null,
                "points": 22,
                "status": "met",
                "details": [],
                "max_points": 22
              },
              {
                "key": "lint_format_config",
                "name": "Lint / format config",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 11
              },
              {
                "key": "static_type_checking",
                "name": "Static type checking",
                "detail": "Rust (statically typed)",
                "points": 11,
                "status": "met",
                "details": [
                  {
                    "code": "statically_typed_language",
                    "params": {
                      "language": "Rust"
                    }
                  }
                ],
                "max_points": 11
              },
              {
                "key": "reproducible_environment",
                "name": "Reproducible environment",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 10
              },
              {
                "key": "demonstrated_agent_practice",
                "name": "Demonstrated agent practice",
                "detail": "100 of the last 100 commits agent-authored or agent-credited",
                "points": 10,
                "status": "met",
                "details": [
                  {
                    "code": "agent_authored_commits",
                    "params": {
                      "count": 100,
                      "sampled": 100
                    }
                  }
                ],
                "max_points": 10
              },
              {
                "key": "automated_maintenance",
                "name": "Automated maintenance",
                "detail": "no automated dependency updates observed",
                "points": 0,
                "status": "missed",
                "details": [
                  {
                    "code": "no_dependency_automation",
                    "params": {}
                  }
                ],
                "max_points": 8
              },
              {
                "key": "openssf_scorecard_pinned_dependencies",
                "name": "OpenSSF Scorecard: Pinned-Dependencies",
                "detail": "dependency not pinned by hash detected -- score normalized to 0",
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 10
              }
            ]
          },
          {
            "key": "ai_code_legibility",
            "band": "excellent",
            "name": "Code legibility for models",
            "note": null,
            "notes": [],
            "value": 97,
            "inputs": {
              "primary_language": "Rust",
              "largest_source_bytes": 375018,
              "source_files_sampled": 85,
              "oversized_source_files": 4
            },
            "components": [
              {
                "key": "type_checkable_code",
                "name": "Type-checkable code",
                "detail": "Rust (statically typed)",
                "points": 45,
                "status": "met",
                "details": [
                  {
                    "code": "statically_typed_language",
                    "params": {
                      "language": "Rust"
                    }
                  }
                ],
                "max_points": 45
              },
              {
                "key": "manageable_file_sizes",
                "name": "Manageable file sizes",
                "detail": "4/85 source files over 60KB",
                "points": 52.4,
                "status": "partial",
                "details": [
                  {
                    "code": "oversized_source_files",
                    "params": {
                      "kb": 60,
                      "sampled": 85,
                      "oversized": 4
                    }
                  }
                ],
                "max_points": 55
              }
            ]
          },
          {
            "key": "ai_interfaces",
            "band": "at_risk",
            "name": "Machine-readable interfaces",
            "note": null,
            "notes": [],
            "value": 40,
            "inputs": {
              "example_dirs": [
                "examples"
              ],
              "has_mcp_signal": false,
              "api_schema_files": []
            },
            "components": [
              {
                "key": "api_schema_openapi_graphql_proto",
                "name": "API schema (OpenAPI/GraphQL/proto)",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 40
              },
              {
                "key": "mcp_server",
                "name": "MCP server",
                "detail": null,
                "points": 0,
                "status": "missed",
                "details": [],
                "max_points": 20
              },
              {
                "key": "runnable_examples",
                "name": "Runnable examples",
                "detail": "examples",
                "points": 40,
                "status": "met",
                "details": [
                  {
                    "code": "file_list",
                    "params": {
                      "files": "examples"
                    }
                  }
                ],
                "max_points": 40
              }
            ]
          }
        ],
        "description": "How well is the repo equipped to be developed and maintained with AI coding agents? An independent, experimental badge — weight 0.0, so it is surfaced on its own and does not affect the overall health score."
      }
    ],
    "metrics_version": "1.13.0"
  },
  "warnings": [
    "Could not fetch crates package 'rusty_h264-bench' from its registry",
    "Could not fetch crates package 'rs-h264-analyzer' from its registry",
    "Could not fetch crates package 'rusty_h264-cli' from its registry",
    "GitHub dependency-graph SBOM unavailable (404); the dependency graph may be disabled for this repository",
    "deps.dev does not index crates:rusty_h264@0.4.0; advisories assessed against the repository dependency graph instead"
  ],
  "report_type": "repository",
  "generated_at": "2026-07-30T10:39:22.876248Z",
  "schema_version": "0.27.0",
  "badge_url": "https://raw.githubusercontent.com/inspect-software/badges/main/v1/r/Remade-With-Rust/rusty_h264.svg",
  "full_name": "Remade-With-Rust/rusty_h264",
  "license_state": "standard",
  "license_spdx": "BSD-2-Clause"
}