公开记录良好
软件健康报告模式 0.27.0 · 指标 1.13.0 · 2026-07-26 15:25 UTC
iree-org / wave
Wave: Python Domain-Specific Language for High Performance Machine Learning
iree-org/wave 的健康指数为 100 分中的 70 分，处于「良好」区间。 其得分最高的类别是Vitality（81/100），最低的是Security（46/100）。 最近一次更新在 26 天前。 近期的大部分工作由 4 位贡献者完成。
70
总分 / 100软件健康指数
指标归入加权类别，统一采用 1–100 量表。总体分先取类别加权平均；当公开证据触发高风险司法辖区政策时，评级会按政策调整，并设置 49（有风险）的上限。AI 就绪度不计入总体分。
70
优秀85-100堪称典范；基本满足所有检验标准
良好70-84健康；仅有轻微不足
中等50-69可接受，但存在明显不足；建议进行审查
存在风险30-49存在重大薄弱环节；采用时应保持审慎
危急1-29问题严重（项目被弃置、仅有单一维护者、缺乏基本工程规范）
评分画像
每条轴代表一个类别。形状比平均值更重要——健康的对象会填满整个图形，而“一峰一谷”式画像意味着某一维度的优势正掩盖另一维度的风险。
所有权
软件包生态系统
| 注册表 | 软件包 | 版本 | 月下载量 | 版本数 | 最近发布 |
|---|---|---|---|---|---|
| PyPI | wave-lang | 3.9.1 | - | 9 | 214 天前 |
按类别列示的指标
活力
项目是否仍有生命——是否仍在编写代码，是否仍在发布版本？
81
开发活跃度
83良好
评分方式
| 28.8/36 | 推送新近度 — 最近一次推送于 26 天前 |
| 27.7/36 | 提交节奏 — 52 周中有 40 周有提交 |
| 18/18 | 提交量 — 最近一年 882 次提交 |
| 0/10 | OpenSSF Scorecard：Maintained — 无数据 |
所用输入
| commits_last_year | 882 |
| human_commit_share | 0.89 |
| days_since_last_push | 26 |
| active_weeks_last_year | 40 |
发布纪律
78良好
评分方式
| 27/27 | 有发布版本 — 已发布 9 个发布版本 |
| 16.2/36 | 发布时效 — 最近一次发布版本于 214 天前 |
| 27/27 | 发布节奏 — 约每 19.7 天发布一次 |
| 0/10 | OpenSSF Scorecard：Signed-Releases — 无数据 |
所用输入
| releases_count | 9 |
| latest_release_tag | v3.9.1 |
| releases_from_tags | 否 |
| days_since_latest_release | 214 |
| mean_days_between_releases | 19.7 |
社区与采用
项目是否拥有用户、下载量与关注度，并具备欢迎贡献者参与的配置？
57
流行度与采用
45存在风险
评分方式
| 28.5/60 | 星标 — 58 个星标 |
| 12.4/25 | 复刻 — 32 个复刻 |
| 3.9/15 | 关注者 — 6 位关注者 |
所用输入
| forks | 32 |
| stars | 58 |
| watchers | 6 |
| growth_state | unverified |
| growth_factor_pct | 100 |
| growth_unverified_reason | no_history |
社区健康
70良好
评分方式
| 22.5/22.5 | README |
| 22.5/22.5 | 许可证 — 可识别的许可证（Apache-2.0） |
| 18/18 | CONTRIBUTING 指南 |
| 0/13.5 | 行为准则 |
| 0/7.2 | 议题模板 |
| 0/6.3 | PR 模板 |
所用输入
| has_readme | 是 |
| has_license | 是 |
| has_contributing | 是 |
| has_issue_template | 否 |
| has_code_of_conduct | 否 |
| has_pull_request_template | 否 |
可持续性与治理
项目能否在其成员之外延续——巴士系数、响应能力、由谁支持，以及软件包的维护状况？
76
维护者韧性（巴士系数）
84良好
评分方式
| 43.2/54 | 巴士系数 — 4 位贡献者贡献了半数提交 |
| 18.6/22.5 | 提交分布 — 头号贡献者编写了 17% 的提交 |
| 13.5/13.5 | 贡献者广度 — 61 位贡献者 |
| 0/10 | OpenSSF Scorecard：Contributors — 无数据 |
所用输入
| bus_factor | 4 |
| contributors_sampled | 61 |
| top_contributor_share | 0.173 |
议题与 PR 响应能力
64中等
评分方式
| 20.7/46.8 | 议题解决 — 44% 的议题已关闭 |
| 33.5/38.3 | PR 接受 — 已裁定的 PR 中 924/1,054 已合并 |
| 0/15 | OpenSSF Scorecard：Code-Review — 无数据 |
所用输入
| merged_prs | 924 |
| open_issues | 93 |
| closed_issues | 74 |
| issue_closed_ratio | 0.443 |
| closed_unmerged_prs | 130 |
所有权与托管责任
66中等
评分方式
| 30/30 | 所有权背书 — 组织持有 |
| 0/20 | 已验证域名 |
| 16.6/25 | 所有者影响力 — iree-org 有 205 位关注者 |
| 19.3/25 | 既往记录 — 33 个公开仓库，账户约 4 年 |
所用输入
| followers | 205 |
| owner_type | Organization |
| is_verified | — |
| owner_login | iree-org |
| public_repos | 33 |
| account_age_days | 1,495 |
软件包维护
91优秀
评分方式
| 25/25 | 已发布且可解析 — pypi 上有 1 个软件包 |
| 26/35 | 发布时效 — 最近一次发布于 214 天前 |
| 20/20 | 版本历史 — 9 个已发布版本 |
| 20/20 | 未被弃用 — 活跃，未被弃用或撤回 |
所用输入
| packages | wave-lang |
| ecosystems | pypi |
| any_deprecated | 否 |
| min_days_since_publish | 214 |
工程质量
基础的工程与文档实践是否到位？
81
工程实践
92优秀
评分方式
| 24/24 | CI 工作流 — 8 个工作流 |
| 24/24 | 存在测试 |
| 16/16 | Linter 配置 |
| 9.6/9.6 | Pre-commit 钩子 |
| 0/6.4 | .editorconfig |
| 0/20 | OpenSSF Scorecard：CI-Tests — 无数据 |
所用输入
| has_ci | 是 |
| has_tests | 是 |
| has_editorconfig | 否 |
| has_linter_config | 是 |
| has_precommit_config | 是 |
文档
65中等
评分方式
| 30/30 | README |
| 25/25 | 文档目录 |
| 0/15 | 文档 / 主页站点 |
| 10/10 | 仓库描述 |
| 0/10 | 主题标签 |
| 0/10 | Wiki |
所用输入
| topics | — |
| has_wiki | 否 |
| homepage | — |
| has_readme | 是 |
| has_docs_dir | 是 |
| has_description | 是 |
安全
可见的安全与供应链实践是否稳固，且不存在未解决的高风险司法辖区暴露？
46
安全态势
33存在风险
评分方式
| 0/30 | 安全策略（SECURITY.md） |
| 25/25 | Dependabot 配置 |
| 0/25 | 依赖锁定文件 — 已发布的类库——锁定文件属于应用层关注点，不作要求 |
| 0/20 | CodeQL 工作流 |
所用输入
| source | file_signals |
| lockfiles | — |
| manifests | pyproject.toml, requirements-iree-pinned.txt, setup.py, water/requirements-dev.txt |
| has_codeql_workflow | 否 |
| has_security_policy | 否 |
| has_dependabot_config | 是 |
已排除计分（无数据或不适用）：依赖锁定文件。 其余权重已重新归一化。
依赖安全公告
100优秀
评分方式
| 35/35 | 直接依赖不含已知公告 — 没有直接依赖携带已知公告 |
| 0/25 | 间接依赖不含已知公告 — 在此范围内，传递依赖集合无法与开发和测试依赖区分 |
| 0/40 | 没有长期未处理的公告 — 没有公告带有发布日期 |
所用输入
| source | osv |
| advisories | 0 |
| affected_packages | 0 |
| assessed_packages | 2 |
| unassessed_packages | 31 |
| affected_by_severity | none |
| direct_affected_packages | 0 |
已排除计分（无数据或不适用）：间接依赖不含已知公告, 没有长期未处理的公告。 其余权重已重新归一化。 已将 2 个已解析依赖与 OSV 比对。 有 31 项无法评估——没有已解析的版本、生态系统不受支持，或超出所报告的软件包清单。 该仓库未发布任何索引可解析的软件包，因此改为评估仓库依赖图。该图将开发与测试版本固定同交付的依赖混在一起，因此仅对声明的运行时依赖计分；传递性发现仅作为背景信息列出，不计入评分。 未对可达性进行分析。
AI 就绪度
该仓库在多大程度上具备与 AI 编码代理协同开发与维护的条件？这是一枚独立的实验性徽章——权重为 0.0，因此单独呈现，不影响总体健康评分。
72
AI 代理上下文与指引
85优秀
评分方式
| 45/45 | 代理指令 — .github/instructions/water-review.instructions.md |
| 0/15 | 机器可读文档（llms.txt） |
| 40/40 | 可读的提交历史 — 89 次人类提交中有 88 次说明了意图（结构化标题或解释性正文） |
所用输入
| has_llms_txt | 否 |
| legible_history_share | 0.989 |
| agent_instruction_files | .github/instructions/water-review.instructions.md |
| agent_instruction_max_bytes | 1,071 |
评分方式
| 12.6/18 | 一条命令的引导启动 — wave_lang/kernel/wave/scheduling/aplp/Cargo.toml（工具链约定，无任务运行器） |
| 22/22 | 自动化测试 |
| 11/11 | Lint / 格式化配置 |
| 0/11 | 静态类型检查 |
| 10/10 | 可复现环境 — devcontainer, Dockerfile |
| 10/10 | 已体现的代理实践 — 最近 100 次提交中有 15 次由代理编写或署名代理 |
| 8/8 | 自动化维护 — 最近 100 次提交中有 1 次为自动依赖更新 |
| 0/10 | OpenSSF Scorecard：Pinned-Dependencies — 无数据 |
所用输入
| has_nix | 否 |
| has_tests | 是 |
| lockfiles | — |
| has_dockerfile | 是 |
| typed_language | 否 |
| bootstrap_files | — |
| has_devcontainer | 是 |
| has_linter_config | 是 |
| typecheck_configs | — |
| agent_commit_share | 0.15 |
| toolchain_manifests | wave_lang/kernel/wave/scheduling/aplp/Cargo.toml |
| dependency_bot_commit_share | 0.01 |
面向模型的代码可读性
53中等
评分方式
| 0/45 | 可类型检查的代码 — Python，未配置类型检查 |
| 53/55 | 可控的文件大小 — 采样的 528 个源文件中有 19 个超过 60KB |
所用输入
| primary_language | Python |
| largest_source_bytes | 162,291 |
| source_files_sampled | 528 |
| oversized_source_files | 19 |
机器可读接口
40存在风险
评分方式
| 0/40 | API 模式（OpenAPI/GraphQL/proto） |
| 0/20 | MCP 服务器 |
| 40/40 | 可运行示例 — examples, notebooks |
所用输入
| example_dirs | examples, notebooks |
| has_mcp_signal | 否 |
| api_schema_files | — |
关键数据
58GitHub 星标
61贡献者
882最近 12 个月提交数
26距最近推送天数
9发布版本数
4巴士系数（bus factor）
93开放议题
PyPI软件包生态系统数
数据采集警告
- Star history unavailable: GitHub GraphQL error: Resource not accessible by personal access token
- OpenSSF Scorecard did not return a usable result (exit code -9); skipping Scorecard checks
更多细节
Star 与 Fork 历史 0 ★ / 32 ⇿
0Star
32Fork
9发布
每颗 star 和每个 fork 的添加时间，来自 GitHub 并按天汇总。累计增长位于其构成来源——每日新增——的正上方，二者可相互对照：稳定的自然增长与短暂的突增形态截然不同。当这一差别可被衡量时，它会作为增长真实性予以报告。
主版本 0次版本 2修订 3
直接依赖 8
| 注册表 | 软件包 | 版本约束 | 清单文件 |
|---|---|---|---|
| PyPI | dill | — | pyproject.toml |
| PyPI | iree-base-compiler | ~=3.12.0rc | pyproject.toml |
| PyPI | iree-base-runtime | ~=3.12.0rc | pyproject.toml |
| PyPI | ml_dtypes | — | pyproject.toml |
| PyPI | numpy | — | pyproject.toml |
| PyPI | sympy | — | pyproject.toml |
| PyPI | torch | >=2.6,<2.10 | pyproject.toml |
| PyPI | typing_extensions | — | pyproject.toml |
全部依赖 33
来自 GitHub 依赖图的完整解析依赖集合：10 个直接依赖与 23 个间接（传递）软件包。仓库提交锁文件时，传递闭包才是完整的。
| 注册表 | 软件包 | 版本 | 关系 |
|---|---|---|---|
| PyPI | dill | — | 直接 |
| PyPI | iree-base-compiler | — | 直接 |
| PyPI | iree-base-compiler | 3.12.0rc20260410 | 直接 |
| PyPI | iree-base-runtime | — | 直接 |
| PyPI | iree-base-runtime | 3.12.0rc20260410 | 直接 |
| PyPI | ml-dtypes | — | 直接 |
| PyPI | numpy | — | 直接 |
| PyPI | sympy | — | 直接 |
| PyPI | torch | — | 直接 |
| PyPI | typing-extensions | — | 直接 |
| crates.io | pyo3 | — | 间接 |
| crates.io | rayon | — | 间接 |
| PyPI | cmake | — | 间接 |
| PyPI | filecheck | — | 间接 |
| PyPI | lit | — | 间接 |
| PyPI | mypy | — | 间接 |
| PyPI | myst-parser | — | 间接 |
| PyPI | nanobind | — | 间接 |
| PyPI | ninja | — | 间接 |
| PyPI | pre-commit | — | 间接 |
| PyPI | pytest | — | 间接 |
| PyPI | pytest-timeout | — | 间接 |
| PyPI | pytest-xdist | — | 间接 |
| PyPI | ruff | — | 间接 |
| PyPI | setuptools | — | 间接 |
| PyPI | setuptools-rust | — | 间接 |
| PyPI | setuptools-scm | — | 间接 |
| PyPI | shibuya | — | 间接 |
| PyPI | sphinx | — | 间接 |
| PyPI | sphinx-autobuild | — | 间接 |
| PyPI | sphinx-rtd-theme | — | 间接 |
| PyPI | sphinxcontrib-mermaid | — | 间接 |
| PyPI | wheel | — | 间接 |
依赖安全公告 0
该仓库未发布可被索引解析的包，因此评估的是其自身的依赖图——共 2 个包，其中也包含从不交付的开发与测试版本固定：0 个存在已知公告，0 个为直接依赖。 有 31 个无法评估——没有已解析的版本、生态系统不受支持，或不在所列包清单之内。
没有已知公告影响已评估的依赖。
公告表示依赖图中记录的版本落入某条公告的受影响范围。可达性未经分析，且依赖图包含开发与测试的版本固定——某项发现可能只涉及工具链而非交付的软件。
原始 JSON 报告 机器可读
{
"data": {
"repo": {
"topics": [],
"is_fork": false,
"size_kb": 14315,
"has_wiki": false,
"homepage": null,
"languages": {
"C": 32028,
"C++": 1853650,
"CSS": 2939,
"MLIR": 1459145,
"Rust": 9239,
"CMake": 35003,
"Shell": 21742,
"Python": 4910136,
"Dockerfile": 654,
"JavaScript": 15608,
"Linker Script": 360
},
"pushed_at": "2026-06-29T15:54:50Z",
"created_at": "2025-07-16T15:01:29Z",
"owner_type": "Organization",
"updated_at": "2026-07-16T15:49:43Z",
"description": "Wave: Python Domain-Specific Language for High Performance Machine Learning",
"is_archived": false,
"is_disabled": false,
"license_spdx": "Apache-2.0",
"default_branch": "main",
"license_spdx_raw": "Apache-2.0",
"primary_language": "Python",
"significant_languages": [
"Python",
"C++",
"MLIR"
]
},
"owner": {
"blog": "iree.dev",
"name": "IREE",
"type": "Organization",
"login": "iree-org",
"company": null,
"location": null,
"followers": 205,
"avatar_url": "https://avatars.githubusercontent.com/u/107954215?v=4",
"created_at": "2022-06-21T19:14:04Z",
"is_verified": null,
"public_repos": 33,
"account_age_days": 1495
},
"license": {
"state": "standard",
"spdx_id": "Apache-2.0",
"raw_spdx": "Apache-2.0",
"file_present": true,
"scorecard_found": null,
"profile_has_license": true
},
"activity": {
"releases": [
{
"tag": "v3.9.1",
"kind": "patch",
"published_at": "2025-12-23T17:45:20Z"
},
{
"tag": "v3.9",
"kind": "other",
"published_at": "2025-11-25T22:34:24Z"
},
{
"tag": "v3.8.2",
"kind": "patch",
"published_at": "2025-11-04T01:01:43Z"
},
{
"tag": "v3.8.1.post1",
"kind": "other",
"published_at": "2025-11-04T00:53:08Z"
},
{
"tag": "v3.8.0",
"kind": "minor",
"published_at": "2025-10-14T18:17:52Z"
},
{
"tag": "v3.7.0",
"kind": "minor",
"published_at": "2025-09-04T23:28:43Z"
},
{
"tag": "v1.0.1",
"kind": "patch",
"published_at": "2025-07-28T22:10:31Z"
},
{
"tag": "v1.0.0-beta.1",
"kind": "prerelease",
"published_at": "2025-07-22T05:08:57Z"
},
{
"tag": "dev-wheels",
"kind": "other",
"published_at": "2025-07-18T18:28:13Z"
}
],
"recent_commits": [
{
"oid": "ceaa2b571b7d4ad10c45978e8adf5c720466e007",
"body": "Diff:\nhttps://github.com/iree-org/iree/compare/iree-3.12.0rc20260409...iree-3.12.0rc20260410\n\nAuto-generated by GitHub Actions using\n[`.github/workflows/update_iree_requirement_pins.yml`](https://github.com/iree-org/wave/blob/main/.github/workflows/update_iree_requirement_pins.yml).\n\nSigned-off-by: \n[…]\n898282+github-actions[bot]@users.noreply.github.com>\nCo-authored-by: iree-pr-automator[bot] <iree-github-actions-bot@google.com>\nCo-authored-by: Gaurav Verma <48321602+xintin@users.noreply.github.com>",
"is_bot": true,
"headline": "Bump IREE requirement pins to their latest versions. (#1289)",
"author_name": "iree-pr-automator[bot]",
"author_login": "iree-pr-automator[bot]",
"committed_at": "2026-04-27T20:50:55Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "f08f0cc073ff63ba3cabf39f76525ee49ceb13d6",
"body": "The core things added are split-k gemm, and it is tested for (1)\ngeneration of the `buffer_atomic_pk_add_bf16` instruction that we wanted\nto use, and (2) for gemm correctness.\n\nOverview of changes unrelated to wave_asm:\n\n- `remove_global_indexing` in `general_utils.py`: Zeroes out tiling\nconstraint \n[…]\n overshoot, avoiding the\nspurious bound. (This was necessary to get scale_preshuffle to have 4x\nvector loads when combined with split-k.)\n\n---------\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "Implement mxfp4 split-k gemm (#958)",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-04-21T15:02:53Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "a69ed25699811d54cdb38727daa3e5a3a4797ad4",
"body": "This PR:\n- Optimizes MXFP 8w schedule with respect to counters and memory ops.\n\n- Includes a handwritten MLIR snippet that performs swizzle and dword\nstores to global memory (instead of u shorts). This optmization brings\napprox 7% improprement.\n\n- Unrolls the kernel twice to remove forced vmcnt(0) +\n[…]\nociated with gather_to_lds operations. This avoids inserting costly\nmasking logic in the dynamic case and improves performance.\n\n\n---------\n\nSigned-off-by: Aurore De Spirlet <aurore.despirlet@amd.com>",
"is_bot": false,
"headline": "MXFP4 8w Optimizations Dynamic (#1211)",
"author_name": "Aurore De Spirlet",
"author_login": "adedespirlet",
"committed_at": "2026-04-17T20:10:44Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "9f2b641fd0081d28ef3d03ac4f4da4debd7e93fb",
"body": "…1292)\n\nAdd interactive IR editing via PassInstrumentation. After (or before) a specified pass, the current IR is written to a temporary file and compilation blocks until the user presses Enter, allowing manual edits to the IR mid-pipeline.\n\n- `--mlir-edit-ir-after=<pass>` / `--mlir-edit-ir-before=<\n[…]\naveCompileOptions` extended with matching fields\n- Lit test covering selective firing, all-pass variants, and IR\nround-trip\n\nFixes #1291\n\n---------\n\nSigned-off-by: Martin Lücke <martin.luecke@amd.com>",
"is_bot": false,
"headline": "[water] Add --mlir-edit-ir-{before,after,all} pass instrumentation (#…",
"author_name": "Martin Paul Lücke",
"author_login": "martin-luecke",
"committed_at": "2026-04-14T07:22:59Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "5966cb13b6b365cac73712a5ef66fe7b6099dbf7",
"body": "As mentioned in PR comments from\nhttps://github.com/iree-org/wave/pull/1266 using `uv` is nice for\ndevelopment generally, so let's just standardize on it in the setup\ninstructions.\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "suggest uv instead of normal pip in docs (#1276)",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-04-13T16:53:07Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "d746da2c7d03f25827fd513bc4c6a77502a6dc3a",
"body": "Use module-level alias for index mappings to decrease their verbosity in\nthe overall IR: only the alias name is printed at use sites and\nidentical mappings have the same alias. This is similar to affine maps.\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>\n\n---------\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>",
"is_bot": false,
"headline": "[water] module-level alias for WaveIndexMappingAttr (#1287)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-10T13:09:29Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "5f41382c6d319848535db3be2e2785dfa54f8683",
"body": "…ison (#1228)\n\nTeach the MLIR-to-FX importer to handle ops used by the MXFP4 GEMM\nkernel and fix the roundtrip comparison for MMA index expressions.\n\nFX emitter:\n- Add `_handle_mma_like_op` handling both `wave.mma` and\n`wave.scaled_mma`\n- Add `_handle_bitcast_op` for `wave.bitcast`.\n- Add derived di\n[…]\nations and lost during MLIR serialization.\n- Add WaveConstraint comparator and thread `subs` into\n`assert_traces_equivalent`\n\nFixes #920\n\n---------\n\nSigned-off-by: Martin Lücke <martin.luecke@amd.com>",
"is_bot": false,
"headline": "[wave2water] Add MXFP4 support to fx_emitter and fix roundtrip compar…",
"author_name": "Martin Paul Lücke",
"author_login": "martin-luecke",
"committed_at": "2026-04-10T12:15:14Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "a3521bc0c63e9fbb07f3608a5b4e38e884adf764",
"body": "## Wide epilogue stores for bf16 MXFP4 GEMM via `v_permlane16_swap`\n\nIn MXFP4 GEMM kernels with bf16 output, the MMA accumulator layout\n(`F32_16x16x128_F8F6F4`) distributes results across threads in a\nnon-contiguous pattern each thread holds 4 values spanning 4 rows but\nonly 1 column. Without optimi\n[…]\n60) | 2936.09 | 3352.67 | 1.142x |\n| (14848, 12672, 10496) | 2781.61 | 3254.58 | 1.170x |\n| (1792, 19968, 60928) | 3115.72 | 3392.02 | 1.089x |\n\n---------\n\nSigned-off-by: xintin <gaurav.verma@amd.com>",
"is_bot": false,
"headline": "Add wide_stores support for MXFP4 (4wave) GEMM (llvm backend) (#1282)",
"author_name": "Gaurav Verma",
"author_login": "xintin",
"committed_at": "2026-04-09T21:05:34Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "518649555bad8f5f04cf8e5cecb3eba096b84ae6",
"body": "This was a clutch introduced to correct for upstream DictionaryAttr not\npreserving the original order of keys when used for index expressions.\nWith WaveSymbolMappingAttr in place, this is no longer necessary and the\ninformation contained in `ordered_syms` is just duplicating what is\navailable in the index expression.\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>\nAssisted-by: GPT 5.3 Codex / Cursor\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>",
"is_bot": false,
"headline": "[water] remove ordered_syms, NFC (#1284)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-09T19:34:38Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "249f7a5d6909f3beeb7cf11ab08052a43f9b376c",
"body": "Diff:\nhttps://github.com/iree-org/iree/compare/iree-3.12.0rc20260408...iree-3.12.0rc20260409\n\nAuto-generated by GitHub Actions using\n[`.github/workflows/update_iree_requirement_pins.yml`](https://github.com/iree-org/wave/blob/main/.github/workflows/update_iree_requirement_pins.yml).\n\nSigned-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>\nCo-authored-by: iree-pr-automator[bot] <iree-github-actions-bot@google.com>",
"is_bot": true,
"headline": "Bump IREE requirement pins to their latest versions. (#1285)",
"author_name": "iree-pr-automator[bot]",
"author_login": "iree-pr-automator[bot]",
"committed_at": "2026-04-09T17:24:44Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "5f6d8bef627466a94843b8db0ba1d5a8dfb66e13",
"body": "Hyperparameters are indexed by symbol names so using a dedicated\ncontainer removes the need to go through opaque attributes or strings.\n\nNote that the definition of the hyperparameter attribute had to be moved\nso it can access the definition of the mapping attribute.\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>\n\n---------\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>",
"is_bot": false,
"headline": "[water] use WaveSymbolMappingAttr for hyperparams (#1279)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-09T16:54:54Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "0ca0dbf5cdd78568e71362750bafa9fa44772309",
"body": "Replace the two parallel ArrayRef parameters (keys/values) with a single\nArrayRef<pair<WaveSymbolAttr, Attribute>> parameter, making the\nattribute directly iterable via range-for. Update all use sites\naccordingly:\n\n- Replace getKeys()/getValues() accessors with structured bindings over\ngetMapping() \n[…]\nhe C API and Python bindings\n\n---------\n\nSigned-off-by: Tim Gymnich <tim@gymni.ch>\nCo-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>\nCo-authored-by: Oleksandr \"Alex\" Zinenko <git@ozinenko.com>",
"is_bot": false,
"headline": "[water] Refactor WaveSymbolMappingAttr to pair-based storage (#1275)",
"author_name": "Tim Gymnich",
"author_login": "tgymnich",
"committed_at": "2026-04-09T08:43:18Z",
"body_truncated": true,
"is_coding_agent": true
},
{
"oid": "6a8d9bbcf488522ec3e7ed0bea37f7467ea5624f",
"body": "LINEAR_INDEX global reads previously forced maskedload\n(linear_buffer_ops = False) to ensure numerics match between IREE and\nwave runtime. However, maskedload on AMD hardware can fault on OOB\naddresses even with an all-false mask, because the hardware may touch\nthe address before checking the mask.\n\n[…]\nes and linearized reads interact\nwith epilogue elimination. So we add one so that we won't accidentally\nbreak epilogue elimination again.\n\n---------\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "reconcile linearization and epilogue elimination (#1278)",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-04-09T05:05:49Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "7e5c23744616708f559a19a7e9bdc05325c82d91",
"body": "This PR moves the scheduling passes to the end of the compilation\npipeline, after all optimization and analysis passes. This lets the\nschedule reason about the final optimized graph : merged reads,\nsimplified indices, computed bounds ,... rather than the intermediate\nstate that existed before.\n\nMovi\n[…]\nory. Fix in update_index_for_unroll() to substitute in dict\nkeys, not just values. And Unroll() to handle bounds expressions too\n\n---------\n\nSigned-off-by: Aurore De Spirlet <aurore.despirlet@amd.com>",
"is_bot": false,
"headline": "Move scheduling passes to the end of the pipeline (#1191)",
"author_name": "Aurore De Spirlet",
"author_login": "adedespirlet",
"committed_at": "2026-04-09T01:36:17Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "dfa45bb3e1a5719ac9c1fd27c40e76e6c2ed8a45",
"body": "This PR adds support for dynamic strides through the waveasm backend.\n\n# 4 Cases for Dynamic Strides\nThere are 4 main cases that needed to be addressed to ensure complete\nsupport.\n\n## Case 1: waveasm + dynamic strides\n\nThe loads were already using correct static offsets. The only issue was\n`num_reco\n[…]\n 1 fix (computeBufferSizeFromMemRef)\n```\n\n---------\n\nSigned-off-by: Surya Jasper <45545431+suryajasper@users.noreply.github.com>\nCo-authored-by: Gaurav Verma <48321602+xintin@users.noreply.github.com>",
"is_bot": false,
"headline": "Dynamic stride support through waveasm (#1091)",
"author_name": "Surya Jasper",
"author_login": "suryajasper",
"committed_at": "2026-04-08T16:17:50Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "5e242938a29a062b7c99ebf5913279fc37a51d38",
"body": "The pinned IREE compiler (3.12.0rc20260407) renamed this flag. Update\nthe two call sites to match.\n\nMade-with: Cursor\n\nSigned-off-by: Aurore De Spirlet <aurore.despirlet@amd.com>",
"is_bot": false,
"headline": "Rename --iree-rocm-target to --iree-hip-target (#1277)",
"author_name": "Aurore De Spirlet",
"author_login": "adedespirlet",
"committed_at": "2026-04-08T16:08:01Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "ce49b1b8ca6642b038de2c7782c28ed7594a9c3e",
"body": "Diff:\nhttps://github.com/iree-org/iree/compare/iree-3.12.0rc20260407...iree-3.12.0rc20260408\n\nAuto-generated by GitHub Actions using\n[`.github/workflows/update_iree_requirement_pins.yml`](https://github.com/iree-org/wave/blob/main/.github/workflows/update_iree_requirement_pins.yml).\n\n---------\n\nSign\n[…]\n Gymnich <tim@gymni.ch>\nCo-authored-by: iree-pr-automator[bot] <iree-github-actions-bot@google.com>\nCo-authored-by: Tim Gymnich <tim@gymni.ch>\nCo-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>",
"is_bot": true,
"headline": "Bump IREE requirement pins to their latest versions. (#1280)",
"author_name": "iree-pr-automator[bot]",
"author_login": "iree-pr-automator[bot]",
"committed_at": "2026-04-08T15:36:17Z",
"body_truncated": true,
"is_coding_agent": true
},
{
"oid": "f559e539a9158d22d9657350d9a8e2a496e16842",
"body": "- Replace identity lambda in applyConstraint with llvm::to_vector\n- Pass getLatticeValue directly instead of wrapping in a lambda\n\nCo-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>\n\nSigned-off-by: Tim Gymnich <tim@gymni.ch>",
"is_bot": false,
"headline": "[water] Simplify map_to_vector lambda expressions in IndexExpr (#1281)",
"author_name": "Tim Gymnich",
"author_login": "tgymnich",
"committed_at": "2026-04-08T14:36:32Z",
"body_truncated": false,
"is_coding_agent": true
},
{
"oid": "ee87b748ded34c5d6ae48984f80f2547be693a13",
"body": "Signed-off-by: Tim Gymnich <tim@gymni.ch>",
"is_bot": false,
"headline": "bump llvm to 6d3d88f74c405db660a20103905787fc67087287 (#1272)",
"author_name": "Tim Gymnich",
"author_login": "tgymnich",
"committed_at": "2026-04-08T12:12:28Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "1ce3ad8e6e681c31e2761b7eb4a5e5b43ab3baf3",
"body": "…(#1266)\n\nThis is based on a build script that I've been using to streamline\nworking with worktrees. IE I wanted a single command to set up a\nworktree (main worktree OR extra worktrees), for myself and for agents\nto easily get going. The main way to use it is to run\n`./build_tools/wave-dev-setup.sh \n[…]\nuirement detection script to\nget the right pytorch. And it always builds waveasm and water.\nBut I think these are all the right defaults.\n\n---------\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "Add wave-dev-setup.sh for worktree-aware build and environment setup …",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-04-07T20:41:00Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "28e8528cc57aa50699e7badf2e41136fb06d89ea",
"body": "Vendor the ixsimpl expression simplifier (commit d41a625) as a\nfirst-party vendored dependency instead of a git+https pin. The full\nproject structure is preserved for easy upstream syncing.\n\nBuild integration: a plain setuptools Extension compiles the amalgamated\nC source into the ixsimpl._ixsimpl C\n[…]\nld_extension now delegates non-CMake extensions to the\nbase class.\n\n---------\n\nSigned-off-by: Ivan Butygin <ivan.butygin@gmail.com>\nCo-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>",
"is_bot": false,
"headline": "Vendor ixsimpl as third_party/ixsimpl (#1217)",
"author_name": "Ivan Butygin",
"author_login": "Hardcode84",
"committed_at": "2026-04-07T18:31:50Z",
"body_truncated": true,
"is_coding_agent": true
},
{
"oid": "140ad78c2f7d615b9e378ac4bbdb9a826189950b",
"body": "Diff:\nhttps://github.com/iree-org/iree/compare/iree-3.12.0rc20260331...iree-3.12.0rc20260407\n\nAuto-generated by GitHub Actions using\n[`.github/workflows/update_iree_requirement_pins.yml`](https://github.com/iree-org/wave/blob/main/.github/workflows/update_iree_requirement_pins.yml).\n\nSigned-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>\nCo-authored-by: iree-pr-automator[bot] <iree-github-actions-bot@google.com>",
"is_bot": true,
"headline": "Bump IREE requirement pins to their latest versions. (#1274)",
"author_name": "iree-pr-automator[bot]",
"author_login": "iree-pr-automator[bot]",
"committed_at": "2026-04-07T17:13:58Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "c04e597770adfa99370c5f2d0ed87c24bd96353b",
"body": "…resentations (#1271)",
"is_bot": false,
"headline": "[wave2water] Teach roundtrip comparison to accept more equivalent rep…",
"author_name": "Martin Paul Lücke",
"author_login": "martin-luecke",
"committed_at": "2026-04-07T15:43:57Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "dc7cdf51bc577541e769d675ca38913daf6c87fe",
"body": "The initial linearization PR didn't apply to GatherToLDS ops. This fixes\nthat and thus reduces register pressure.\n\nThis enables annotate_iv_strides (which already handles GatherToLDS) and\nthe codegen LINEAR_INDEX path to work for GatherToLDS ops.\n\n---------\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "Extend flatten_read_indices to linearize GatherToLDS src indices (#1241)",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-04-07T15:11:54Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "16ffa6140d9724a4417cf40203839d23839e0341",
"body": "Diff:\nhttps://github.com/iree-org/iree/compare/iree-3.12.0rc20260330...iree-3.12.0rc20260331\n\nAuto-generated by GitHub Actions using\n[`.github/workflows/update_iree_requirement_pins.yml`](https://github.com/iree-org/wave/blob/main/.github/workflows/update_iree_requirement_pins.yml).\n\nSigned-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>\nCo-authored-by: iree-pr-automator[bot] <iree-github-actions-bot@google.com>",
"is_bot": true,
"headline": "Bump IREE requirement pins to their latest versions. (#1214)",
"author_name": "iree-pr-automator[bot]",
"author_login": "iree-pr-automator[bot]",
"committed_at": "2026-04-07T13:48:04Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "4ee4b11b703c9d02bd80a8c9a7217e54050f3379",
"body": "Avoids most string attribute boxing/unboxing.",
"is_bot": false,
"headline": "[water] use WaveSymbolMappingAttr in lattice consistently (#1270)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-07T12:43:50Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "47516d49bfe128d247afbb37f7f35b65989736cf",
"body": "Unlike DictionaryAttr, this container preserves the order of keys given\nat construction (as opposed to sorting alphabetically), which is\nload-bearing for later stages of Wave compilation.\n\nThis also avoids the reliance on strings and string comparisons when\nsymbols are processed.",
"is_bot": false,
"headline": "[water] use WaveSymbolMappingAttr for index expressions (#1235)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-07T11:15:03Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "58d97282b2d84d4b2365f23f45de8bf78d5bf675",
"body": "…(#1269)\n\nThe ceiling() wrapping that handles bare fractions before affine conversion was gated on `isinstance(expr, sympy.Mul)`. Compound expressions like `A + B*floor(T/64)/2` (from MXFP4 wave_shape=(2,2)\ndivision) have `Add` as the top-level type and bypassed the guard, hitting `AffineConversionE\n[…]\nconditionally. `sympy.ceiling()` distributes correctly, and the converter's existing `ceiling` handler produces valid `ceildiv` affine expressions.\n\nSigned-off-by: Martin Lücke <martin.luecke@amd.com>",
"is_bot": false,
"headline": "[water_emitter] Fix ceiling wrapping for compound affine expressions …",
"author_name": "Martin Paul Lücke",
"author_login": "martin-luecke",
"committed_at": "2026-04-07T11:03:56Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "c2d2e7b052cfae841d081041690c20759697823c",
"body": "Bumps the github-actions group with 7 updates in the / directory:\n\n| Package | From | To |\n| --- | --- | --- |\n| [pypa/cibuildwheel](https://github.com/pypa/cibuildwheel) | `3.4.0` |\n`3.4.1` |\n| [actions/upload-artifact](https://github.com/actions/upload-artifact)\n| `4.6.2` | `7.0.0` |\n|\n[actions/do\n[…]\n of the specified dependency and ignore\nconditions\n\n\n</details>\n\nSigned-off-by: dependabot[bot] <support@github.com>\nCo-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>",
"is_bot": true,
"headline": "Bump the github-actions group across 1 directory with 7 updates (#1264)",
"author_name": "dependabot[bot]",
"author_login": "dependabot[bot]",
"committed_at": "2026-04-07T10:52:22Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "1c9d0c1ddb5da0397d274e738d91b925e307531f",
"body": "The major blocker for universal read linearization was non-contiguous\ntensors under the LLVM dynamic-stride ABI. This PR enables\nflatten_read_indices for those paths by default and guards correctness\nwith a runtime contiguity assertion at kernel launch.\n\nWhat changes:\n\n- flatten_read_indices now run\n[…]\nts, import ordering fix.\n\nThis does not yet remove the linearize_reads compilation option. We\nstill need Water support and a few other minor things.\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "Enable linearization for dynamic-stride and attention paths (#1240)",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-04-06T21:56:40Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "59902dd116b3f788ca28f9e0cf0f4c324bcabf8a",
"body": "…dedup, soffset folding, and SGPR address chain support (#1242)\n\nAdd localCSERange for cloned address chain deduplication. Add voffset\ndecomposition into (VGPR base, SGPR addend, instOffset constant) for\ncross-load deduplication. Add peepholeSoffsetFold to fold constant\nsoffsets into instOffset. Sup\n[…]\nrom gfx942 to gfx950\n- Added negative stride lit test (downward-counting IV with step=-1)\n\n---------\n\nSigned-off-by: xintin <gaurav.verma@amd.com>\nCo-authored-by: Sanket Pandit <sanket.pandit@amd.com>",
"is_bot": false,
"headline": "[asm] Enhance buffer load strength reduction with local CSE, voffset …",
"author_name": "Gaurav Verma",
"author_login": "xintin",
"committed_at": "2026-04-04T05:32:01Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "eae862b9ad0b09adfebc20f69667601bcaaf5a18",
"body": "…gmentation (#1236)\n\nPost-allocation pass that reassigns physical VGPRs using shortest-first\ngreedy packing to eliminate fragmentation from interleaved `buffer_load`\n(long-lived) and `ds_read` (short-lived) destinations. Wired as\n`--waveasm-vgpr-compaction` in the pass pipeline after linear scan\nall\n[…]\nady locally optimal for minimizing `.vgpr_count`; the\npost-pass closes the remaining alignment gaps.\n\nSigned-off-by: xintin <gaurav.verma@amd.com>\nCo-authored-by: Sanket Pandit <sanket.pandit@amd.com>",
"is_bot": false,
"headline": "[asm] Cherry-pick: Add VGPR compaction pass to eliminate register fra…",
"author_name": "Gaurav Verma",
"author_login": "xintin",
"committed_at": "2026-04-03T23:12:03Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "6f4bd59b2eb35da77d850bfd3480fe371a0ddfa8",
"body": "Add VGPR/SGPR rematerialization to shorten live ranges before allocation\nCherry-picked from\n[panditsa/wave@49d096c974c3](https://github.com/iree-org/wave/commit/49d096c974c3e7c99dddb9e5e63628329c2348a6)\n\n([4waveasm-256x192x256](https://github.com/iree-org/wave/compare/main...panditsa:wave:4waveasm-a\n[…]\nrity\n - Add pressure-aware TODO replacing vague \"sophisticated algorithm\"\n - Add lit tests covering all three remat patterns + loop guard\n\n---------\n\nSigned-off-by: xintin <gaurav.verma@amd.com>",
"is_bot": false,
"headline": "[asm]cherry pick and add tests for s/vgpr remat changes (#1226)",
"author_name": "Gaurav Verma",
"author_login": "xintin",
"committed_at": "2026-04-03T21:51:07Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "397a29833d48e9e45fc07a7f4e1f8eabafa893ef",
"body": "Model the AMDGCN Scalar Condition Code (SCC) flag as a first-class MLIR\ntype in WaveASM instead of treating it as an implicit side effect.\n\n### Problem\n\nMany SALU instructions (bitwise, shifts, adds, compares) clobber the\nhardware SCC flag, but WaveASM previously had no way to represent this\nin the \n[…]\nllision with another PR\n\nSigned-off-by: Ivan Butygin <ivan.butygin@gmail.com>\nCo-authored-by: Sanket Pandit <sanket.pandit@amd.com>\nCo-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>",
"is_bot": false,
"headline": "Reland [asm] Explicit SCC modeling (#1219) (#1239)",
"author_name": "Ivan Butygin",
"author_login": "Hardcode84",
"committed_at": "2026-04-03T20:55:48Z",
"body_truncated": true,
"is_coding_agent": true
},
{
"oid": "d8b7c88f1620226f2c16c6ba241899ae9689f8a6",
"body": "This reverts commit b07277d3b05989d004fa87960cd2af4408c96364.\n\nSigned-off-by: xintin <gaurav.verma@amd.com>",
"is_bot": false,
"headline": "Revert \"[asm] Explicit SCC modeling (#1219)\" (#1237)",
"author_name": "Gaurav Verma",
"author_login": "xintin",
"committed_at": "2026-04-03T19:37:34Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "d99e86e3160162e97358099d5687e1ced6a12886",
"body": "…ation (#1218)\n\n- **Replace O(n) RegPool with O(1) BitVector and add bidirectional VGPR\nallocation** — Register operations use `llvm::BitVector` instead of\nsorted free-list scans. Multi-register VGPR ranges spanning >75% of the\nprogram are allocated from the top of expected usage, separating\nlong-li\n[…]\nw `linear-scan-bidirectional.mlir` covers long-lived\nvs short-lived allocation, loop double-buffer pattern, and\nsingle-reg/SGPR bypass.\n\n---------\n\nSigned-off-by: Sanket Pandit <sanket.pandit@amd.com>",
"is_bot": false,
"headline": "[asm] Rewrite RegPool with BitVector and add bidirectional VGPR alloc…",
"author_name": "Sanket Pandit",
"author_login": "panditsa",
"committed_at": "2026-04-03T17:37:07Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "b319e9161bb17f01c8f193a5142c6d7cb17cc6cd",
"body": "…LU add/mul (#1225)\n\n- **Add compile-time constant folding for `V_ADD_U32` and\n`V_MUL_LO_U32`** — When both operands are compile-time constants\n(including through `V_MOV_B32` indirections via `getConstantValue`),\nfold the operation into a `v_mov_b32` of the computed result with 32-bit\nwrapping. Mult\n[…]\ncovering constant folding (add, mul), 32-bit\noverflow wrapping, large-operand truncation, and MOV-indirection for\nzero/one elimination.\n\n---------\n\nSigned-off-by: Sanket Pandit <sanket.pandit@amd.com>",
"is_bot": false,
"headline": "[asm] Enhance peephole pass with compile-time constant folding for VA…",
"author_name": "Sanket Pandit",
"author_login": "panditsa",
"committed_at": "2026-04-03T17:00:56Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "b07277d3b05989d004fa87960cd2af4408c96364",
"body": "Model the AMDGCN Scalar Condition Code (SCC) flag as a first-class MLIR\ntype in WaveASM instead of treating it as an implicit side effect.\n\n### Problem\n\nMany SALU instructions (bitwise, shifts, adds, compares) clobber the\nhardware SCC flag, but WaveASM previously had no way to represent this\nin the \n[…]\nte operands.\n\n---------\n\nSigned-off-by: Ivan Butygin <ivan.butygin@gmail.com>\nCo-authored-by: Sanket Pandit <sanket.pandit@amd.com>\nCo-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>",
"is_bot": false,
"headline": "[asm] Explicit SCC modeling (#1219)",
"author_name": "Ivan Butygin",
"author_login": "Hardcode84",
"committed_at": "2026-04-03T16:40:03Z",
"body_truncated": true,
"is_coding_agent": true
},
{
"oid": "a5f720d2229f1cc33eb547ee0e38edf167d8c671",
"body": "…ss sort order + tests and fixes (#1220)\n\n- Cherry-picks\n[125cdc2](https://github.com/iree-org/wave/commit/125cdc213593411768ea494088e2f12eb81849c7)\n(\"Generalize WAR hazard detection and improve liveness sort order\")\nwhich extends WAR hazard detection beyond buffer_load to cover any\niter_arg/block_a\n[…]\nalue always has a defining op\n> - Adds tie-break rationale to the sort comparator comment\n\n---------\n\nSigned-off-by: xintin <gaurav.verma@amd.com>\nCo-authored-by: Sanket Pandit <sanket.pandit@amd.com>",
"is_bot": false,
"headline": "[asm] Cherry pick: Generalize WAR hazard detection and improve livene…",
"author_name": "Gaurav Verma",
"author_login": "xintin",
"committed_at": "2026-04-03T16:10:12Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "059c25ba2c0ee76bd260f3ead5e92dfee1407c25",
"body": "The main goal of this is to be able to have memory addresses for reads\nin a loop be simplified to `start + IV * stride`, with each of those\nvalues either being constant or at least able to be hoisted outside of\nthe loop body.\n\nAdds a pre-codegen pipeline that flattens N-dimensional read addresses\nin\n[…]\n reads\n\nThis adds some new lit tests that show that with our mxfp4 shuffle\nlayout we can generate linearized reads with constant stride.\n\n\n---------\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "Linearize read indices pipeline (#1196)",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-04-03T15:06:46Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "e85de690aa6c77fcef9eb9fd610a38f0ff98410b",
"body": "ScaledMMA is an MMA for heuristic purposes, should be treated in the\nsame way.\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>",
"is_bot": false,
"headline": "[water] account for ScaledMMA in propagation heuristic (#1234)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-03T14:20:03Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "e2e5faa6113eddf2a6219d8409f3a09a6f140607",
"body": "Mostly NFC.\n\nThis provides a more sound alternative with the right key type to\nDictionaryAttr and will allow to remove string-based comparisons.\n\n---------\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>",
"is_bot": false,
"headline": "[water] use WaveSymbolMappingAttr for vector shapes (#1233)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-03T13:04:19Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "bef40dba3767287b6a0b9c77bbff14d33c9fc14d",
"body": "Group all logic pertaining to index expressions in\nIndexExprInference.h/cpp. This fixes the layering between interfaces\nthat need be op-agnostic and analysis using the interface that isn't.\n\nMove the index expression inference pass into its own file from\nInferTypes.cpp where it was originally bootstrapped to reuse lattice\nutilities.\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>\nCo-authored-by: Claude Opus 4.6 <noreply@anthropic.com>",
"is_bot": false,
"headline": "[water] reorganize index expr related files, NFC (#1231)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-03T08:55:09Z",
"body_truncated": false,
"is_coding_agent": true
},
{
"oid": "3a3fd6fcdb797dd6228ccccb047eb929e06e08ac",
"body": "This is a separate field of the index expression lattice so it is\nhelpful to highlight which field is causing conflicts.\n\n---------\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>\nCo-authored-by: Tim Gymnich <tim@gymni.ch>",
"is_bot": false,
"headline": "[water] improve error message for source vector shape conflict (#1230)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-02T21:18:28Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "baaff6cfbc2231040fc9c23174ba428057b84515",
"body": null,
"is_bot": false,
"headline": "Add a e2e test for water index exprsssions in attention (#1171)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-02T19:18:27Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "193d30b02c6bf35ea891d0f5c85020594c5ff407",
"body": "Signed-off-by: Alex Zinenko <git@ozinenko.com>",
"is_bot": false,
"headline": "fix syntax error (#1232)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-02T16:54:39Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "de592dcd5df783b78c8c39e1304484a854e3a1fb",
"body": "It is unfortunately necessary to introduce this additional complexity to\nreflect the naive non-dataflow approach in the python counterpart: keep\ntrack of the vector shape of the operation from which the lattice\noriginally propagates. This is required by the propagation-stopper\nheuristic that cannot \n[…]\ne, but necessary to match\nthe current behavior. It must be revised as soon as possible to avoid\npriorities and non-local behavior altogether.\n\n---------\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>",
"is_bot": false,
"headline": "[water] provenance-based propagation heuristic (#1206)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-02T15:41:34Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "d0c4b8bd0c21fc3c8cabcc1738413bf3f45e295c",
"body": "This has been failing intermittently on CI for over a month with no\nintention to fix, I'm tired of having to triage.\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>",
"is_bot": false,
"headline": "disable flaky rdna4 test (#1229)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-04-02T15:38:24Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "c283f993f95d6fb58075f1515aeaebdeab65cc42",
"body": "- Add ScaledMMAOp\nfixes #924\n\n---------\n\nSigned-off-by: Tim Gymnich <tim@gymni.ch>",
"is_bot": false,
"headline": "[water] Add ScaledMMAOp (#1136)",
"author_name": "Tim Gymnich",
"author_login": "tgymnich",
"committed_at": "2026-04-02T13:50:40Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "b2301899ffb626bf4cd269451816f02690358f04",
"body": "Add a progressive MLIR roundtrip test for the tagged MXFP4 GEMM kernel\n(`get_tagged_mxfp4_gemm` + `get_mxfp4_dbuf_schedule`), matching the\nconfiguration used in `examples/python/7.1_schedule.py`. All 49 pipeline\npasses are xfailed, with currently visible causes being:\n\n- Passes 1-11: emitter lacks a\n[…]\nn_progressive_roundtrip` resilient to pass crashes\nvia try/except/else so a single crashing pass no longer aborts the\nremaining passes.\n\nFixes #925\n\nSigned-off-by: Martin Lücke <martin.luecke@amd.com>",
"is_bot": false,
"headline": "[wave2water] Add MXFP4 GEMM progressive roundtrip test (#1224)",
"author_name": "Martin Paul Lücke",
"author_login": "martin-luecke",
"committed_at": "2026-04-02T06:48:44Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "6fb475195ff2eaea8e8cf932a96307010394d92d",
"body": "…209)\n\nAdd NonEmittingOp trait to replace the hardcoded op list in\nHazardMitigation, and add InsertOp as the dual of ExtractOp for in-place\nsub-register replacement in wide registers.\n\nNonEmittingOp trait is applied to PackOp, ExtractOp, InsertOp,\nPrecoloredSRegOp, PrecoloredVRegOp, ConstantOp, DCEP\n[…]\nandleFatRawBufferCast\nsuppress-swizzle path for PackOp-built SRDs.\n\n---------\n\nSigned-off-by: Ivan Butygin <ivan.butygin@gmail.com>\nCo-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>",
"is_bot": false,
"headline": "[waveasm] Add NonEmittingOp trait and InsertOp to WaveASM dialect (#1…",
"author_name": "Ivan Butygin",
"author_login": "Hardcode84",
"committed_at": "2026-04-01T15:17:58Z",
"body_truncated": true,
"is_coding_agent": true
},
{
"oid": "725f3bae7c4819c76cba95f7a72ade2ebcf790f7",
"body": "…R roundtrip (#1213)\n\nFix `in_thread_transpose` passing a bare `int` to `Reshape.target_vector_shape` where the type annotation requires `dict[IndexSymbol, int]`. This was the only call site violating the contract (e.g. `decompose_scan_ops.py` does it correctly). Removes the FIXME(#873) workaround i\n[…]\nr always recreates one per iterate result.\n\nTogether, these fixes resolve all outstanding round-trip failures for the attention kernel.\nFixes #1019\n\nSigned-off-by: Martin Lücke <martin.luecke@amd.com>",
"is_bot": false,
"headline": "[wave2water] Fix Reshape type violation and enable full attention MLI…",
"author_name": "Martin Paul Lücke",
"author_login": "martin-luecke",
"committed_at": "2026-04-01T09:29:19Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "99697f0bcd64898455ddf720732f0d414df3c7c9",
"body": "When the compiler emits IR for sympy expressions containing multiple\nfraction additions, it defaulted to cross-multiplication even when the\ndenominator was the same across all fractions. This caused denominator\nexplosion. This is now fixed : it identifies when denominators are\nidentical and directly adds the corresponding numerators instead.\n\n---------\n\nSigned-off-by: Aurore De Spirlet <aurore.despirlet@amd.com>",
"is_bot": false,
"headline": "Optimize same-denominator fraction addition in gen_sympy_index (#1212)",
"author_name": "Aurore De Spirlet",
"author_login": "adedespirlet",
"committed_at": "2026-04-01T00:32:34Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "0d4eb3f916ba7e82e1fb6d9729246af99292e420",
"body": "Fix opsel scaled_mfma pass for (2,2) wave shapes and asymmetric blocks\n\nSquashed cherry-pick of 7 opsel/scale coalescing fixes from\n`panditsa:4waveasm-256x192x256`:\n1. Handle untraceable yield values in _find_mergeable_groups (aa2a9754)\n2. Add _merge_scale_byte_loads for unrolled sub-word scale load\n[…]\n was passed to scf.YieldOp, which crashed.\nFix: Added a filter to drop unresolved groups (yield_source=None) before\nbuilding the coalesce plan.\n\n---------\n\nSigned-off-by: xintin <gaurav.verma@amd.com>",
"is_bot": false,
"headline": "[asm] Cherry pick opsel scaled mfma fix (#1215)",
"author_name": "Gaurav Verma",
"author_login": "xintin",
"committed_at": "2026-03-31T22:59:21Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "0cbff5e8406989e3e10753532820c4917a815a8d",
"body": "Support roundtripping vector shapes. This required special handling for\nMMA ops where, on the MLIR side, we set vector shapes as a list for LHS,\nRHS, Accumulator and Result similarly to index expressions and, on the\nFX side, there is a single vector shapes entry that contains dimensions\nfrom all val\n[…]\nrateOp in the dialect\ninstead since GetResult are fundamentally absent from the MLIR\ncounterpart. An alternative solution could be to simply ignore GetResult\nduring comparison of vector shape entries.",
"is_bot": false,
"headline": "[wave2water] roundtrip vector shapes (#1202)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-31T10:41:44Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "46a069d90204092866a0e74f0e428b9a72658d13",
"body": "- Adds `_resolve_chained_subs` to\n`wave_lang/kernel/_support/indexing.py` which topologically sorts\nsubstitution entries by their symbol dependencies, ensuring derived\nsymbols like `K_SCALE = K // 32` fully resolve when `K` is also in the\nsubstitution dict.\n- `IndexingContext.set_subs` now calls `_r\n[…]\ngle/multi-level chains, floor-div,\nPiecewise, circular dependency warnings, order independence, and\nintegration with `IndexingContext`.\n\n---------\n\nSigned-off-by: Sanket Pandit <sanket.pandit@amd.com>",
"is_bot": false,
"headline": "Resolve chained symbol dependencies in IndexingContext.set_subs (#1183)",
"author_name": "Sanket Pandit",
"author_login": "panditsa",
"committed_at": "2026-03-30T21:29:40Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "f2028ec55c8e61bcacbc22232b1f596609d39e9e",
"body": "Diff:\nhttps://github.com/iree-org/iree/compare/iree-3.12.0rc20260327...iree-3.12.0rc20260330\n\nAuto-generated by GitHub Actions using\n[`.github/workflows/update_iree_requirement_pins.yml`](https://github.com/iree-org/wave/blob/main/.github/workflows/update_iree_requirement_pins.yml).\n\nSigned-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>\nCo-authored-by: iree-pr-automator[bot] <iree-github-actions-bot@google.com>",
"is_bot": true,
"headline": "Bump IREE requirement pins to their latest versions. (#1207)",
"author_name": "iree-pr-automator[bot]",
"author_login": "iree-pr-automator[bot]",
"committed_at": "2026-03-30T20:24:52Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "e94fc5a59ee4975d815a59f6bc459c95fc9ef86f",
"body": "…(#1188)\n\nEliminate all raw assembly string manipulation from SRD (Shader Resource\nDescriptor) construction, replacing it with typed MLIR ops visible to\nthe register allocator.\n\n### What changed\n\n**Core refactor:** SRD construction in `handleFatRawBufferCast`,\n`emitSRDBaseAdjustment`, and `handleMak\n[…]\n pass now inserts explicit copies for ExtractOp\ninputs to PackOps.\n\n---------\n\nSigned-off-by: Ivan Butygin <ivan.butygin@gmail.com>\nCo-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>",
"is_bot": false,
"headline": "[asm] Replace RawOp with typed SALU ops + PackOp in SRD construction …",
"author_name": "Ivan Butygin",
"author_login": "Hardcode84",
"committed_at": "2026-03-30T19:02:53Z",
"body_truncated": true,
"is_coding_agent": true
},
{
"oid": "d518d4ef77c5f9234ca2b2f458b433b50534b25a",
"body": "…1184)\n\nisMultiWaveKernel() only checked workgroup_size_y > 1 or _z > 1, missing\nkernels like [256, 1, 1] that pack 4 waves into the X dimension. This\ncaused gpu.thread_id to emit v_mbcnt (lane_id 0-63) instead of reading\nhardware v0 (workitem_id_x 0-255), so every wave saw the same thread IDs\nand wrote to the same output region.\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "Fix isMultiWaveKernel() to detect multi-wave kernels with wgX > 64 (#…",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-03-30T17:12:35Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "52f520409118b4fe09f4ac85ff609665c2420edb",
"body": "…calization (#1208)\n\nPreviously, `vector_shapes` were only present on `IterArg` and\n`GetResult` nodes from the original graph, but not propagated when new\ninstances of these nodes were created during software pipelining or when\ncapture placeholders were materialized for region-captured operands.\n\nTh\n[…]\nlders created by\n`materialize_capture_placeholder` and `_canonicalize_nested_region`,\nfixing missing shapes on region-captured operands.\n\n---------\n\nSigned-off-by: Martin Lücke <martin.luecke@amd.com>",
"is_bot": false,
"headline": "Propagate vector_shapes through loop reconstruction and region canoni…",
"author_name": "Martin Paul Lücke",
"author_login": "martin-luecke",
"committed_at": "2026-03-30T16:18:43Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "6e3416cc79cf95b1255390ca62ff8c5334f62ec9",
"body": "…(#1197)\n\nWhen the waveasm backend runs out of registers, the assembler previously\nproduced many repeated \"register index is out of range\" errors (one per\ninstruction using an out-of-range register). Replace this with a single\ndiagnostic that states how many registers the kernel needs vs. how many\na\n[…]\nlimit (which can happen when\nthe emitter adds swap-temp registers for loop back-edge copies), emit\none error before writing any assembly.\n\n---------\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "[waveasm] Emit a single clear error when register budget is exceeded …",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-03-30T16:06:18Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "35799f1b4410a2ff1a3c934e6743b6280e3ba524",
"body": "Diff:\nhttps://github.com/iree-org/iree/compare/iree-3.12.0rc20260325...iree-3.12.0rc20260327\n\nAuto-generated by GitHub Actions using\n[`.github/workflows/update_iree_requirement_pins.yml`](https://github.com/iree-org/wave/blob/main/.github/workflows/update_iree_requirement_pins.yml).\n\n---------\n\nSign\n[…]\nreply.github.com>\nSigned-off-by: Ivan Butygin <ivan.butygin@gmail.com>\nCo-authored-by: iree-pr-automator[bot] <iree-github-actions-bot@google.com>\nCo-authored-by: Ivan Butygin <ivan.butygin@gmail.com>",
"is_bot": true,
"headline": "Bump IREE requirement pins to their latest versions. (#1201)",
"author_name": "iree-pr-automator[bot]",
"author_login": "iree-pr-automator[bot]",
"committed_at": "2026-03-28T23:52:11Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "2989d7a056493439c64d5c9d66572b768afcf314",
"body": "Introduces wave.bitcast, a new Water dialect op that reinterprets the\nraw bits of a register-resident tensor as a different element type.\nUnlike a plain type cast, bitcast preserves the total bit count by\nscaling exactly one tensor dimension by the source-to-destination\nelement bitwidth ratio.\n\nNew \n[…]\n dimension symbolically.\n\nLowering\n- Pattern to lower wave.bitcast between fully-typed wave tensors to an\narith.bitcast on the underlying vector.\n\nfixes #922\n\nSigned-off-by: Tim Gymnich <tim@gymni.ch>",
"is_bot": false,
"headline": "[water] Add BitcastOp and Scaled Dimensions (#1135)",
"author_name": "Tim Gymnich",
"author_login": "tgymnich",
"committed_at": "2026-03-28T08:38:06Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "2d2aeaa16ae0e97cfa86b13e41d62c9ee9926fd9",
"body": "Reverts iree-org/wave#1187\n\nThis is breaking some tests, but wasn't caught by CI because of a race\nin merges.",
"is_bot": false,
"headline": "Revert \"Bump IREE requirement pins to their latest versions.\" (#1204)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-27T17:11:45Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "cbbca6529bec51531577e7009b6e1971aee0fe23",
"body": "Diff:\nhttps://github.com/iree-org/iree/compare/iree-3.12.0rc20260324...iree-3.12.0rc20260325\n\nAuto-generated by GitHub Actions using\n[`.github/workflows/update_iree_requirement_pins.yml`](https://github.com/iree-org/wave/blob/main/.github/workflows/update_iree_requirement_pins.yml).\n\nSigned-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>\nCo-authored-by: iree-pr-automator[bot] <iree-github-actions-bot@google.com>",
"is_bot": true,
"headline": "Bump IREE requirement pins to their latest versions. (#1187)",
"author_name": "iree-pr-automator[bot]",
"author_login": "iree-pr-automator[bot]",
"committed_at": "2026-03-27T09:08:30Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "e3e3cd9be5800246c41ca49f0e3f414f43932484",
"body": "Port the heuristic stopping index expression propagation from pywave. It\nis vaguely based on the notion of completeness and avoids propagating\nincomplete index expression sets, i.e., sets that are not covering\ndimensions that will be subject to expansion later on. In practice, it\nserves to avoid expression conflicts in attention kernels.",
"is_bot": false,
"headline": "[water] index propagation stopper heuristic (#1165)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-27T09:06:22Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "775f1a3d5b239e09b46b091b16bdddfa8a14021e",
"body": "All of our requirements files have a maximum bound of torch <2.9, except\nour pyproject.toml was different (<2.10). This standardizes them by\nbumping up to 2.10 everywhere. Previously a bump up broke CI, but I want\nto try again without the other changes that were in the last PR that did\nthis, and see whether they are real failures or just flakey tests.\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "standardize torch version bounds, bumping up to 2.10 (#1190)",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-03-26T20:44:04Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "4c832971c2b78c182cee163054ceaf6493e27a0b",
"body": "Signed-off-by: Tim Gymnich <tim@gymni.ch>\nCo-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>",
"is_bot": false,
"headline": "Add requirements-pytorch-rocm-generated.txt to gitignore (#1195)",
"author_name": "Tim Gymnich",
"author_login": "tgymnich",
"committed_at": "2026-03-26T12:41:21Z",
"body_truncated": false,
"is_coding_agent": true
},
{
"oid": "f8519c917f5a01974cc91dc35f64102fbd42e921",
"body": "remove left over write to file\n\n🤖 Generated with [Claude Code](https://claude.com/claude-code)\n\nSigned-off-by: Tim Gymnich <tim@gymni.ch>\nCo-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>",
"is_bot": false,
"headline": "Remove debug output in tests (#1193)",
"author_name": "Tim Gymnich",
"author_login": "tgymnich",
"committed_at": "2026-03-26T11:58:25Z",
"body_truncated": false,
"is_coding_agent": true
},
{
"oid": "32e5fc752d125431b6812f7b8c59badc250029ed",
"body": "This is similar to index expressions and should arguably be integrated\nwith them. For now, attaching these as attributes for consistency with\npywave.",
"is_bot": false,
"headline": "[water] attach vector shapes to ops (#1181)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-26T10:46:07Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "951ad78ddcfc2ceea4bbf5a219255422066f04fc",
"body": "## Summary\n- Bumps `iree-base-compiler` and `iree-base-runtime` from `~=3.11.0rc`\nto `~=3.12.0rc` in `pyproject.toml`\n\n🤖 Generated with [Claude Code](https://claude.com/claude-code)\n\nSigned-off-by: Tim Gymnich <tim@gymni.ch>\nCo-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>",
"is_bot": false,
"headline": "Bump IREE RC version to 3.12 (#1192)",
"author_name": "Tim Gymnich",
"author_login": "tgymnich",
"committed_at": "2026-03-26T09:16:06Z",
"body_truncated": false,
"is_coding_agent": true
},
{
"oid": "396ee770f9286faee162ebe6fba99d5d2301a8e8",
"body": "The gen-pytorch-rocm-requirements.py script was selecting the first\nwheel source whose URL responded to an HTTP HEAD, without checking\nwhether it actually contained a torch version matching the spec. When\nthe PyTorch Foundation rocm7.2 index only carries torch>=2.11 but\nTORCH_SPEC requires <2.9, the\n[…]\ning a source.\nThis lets the script correctly fall through to the AMD --find-links repo\nwhen the PyTorch index lacks a compatible version.\n\n---------\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "Probe for matching torch version, not just live URL (#1185)",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-03-25T15:53:05Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "ca6e6fd112d65b4ea36f5f759c95f2befec38196",
"body": "Only add them to tensors indexed by the induction variable.",
"is_bot": false,
"headline": "[water] don't add tiling constraints to irrelevant indexes (#1170)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-25T12:00:33Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "d4ddd3d9a73f509e1731308c55cf1f7879ccdc28",
"body": "Remove the microoptimization since it results in `getresult` operation\nhaving index expressions, just incorrect ones, confusing for whoever\nattempts to debug the algorithm.",
"is_bot": false,
"headline": "propagate indices into getresult (#1169)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-25T12:00:10Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "65b65882849add20b5e41f0cf434f7310b2bc7ab",
"body": "Add instructions attempting to improve code review quality.",
"is_bot": false,
"headline": "[water] add copilot instructions (#1186)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-25T11:59:32Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "ee4018ac38ce72f16e4f774efd9588d44cbba808",
"body": "These were not supported originally but may happen in practice, in\nparticular for fused attentions.",
"is_bot": false,
"headline": "[wave2water] water analysis for multi-result iterates (#1167)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-25T09:27:21Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "7e85ef37c627fb6cede30af4451ff2fd103b8ed8",
"body": "Identification of dimensions categories wasn't taking into account batch\ndimensions since it was written before we introduced those.",
"is_bot": false,
"headline": "[wave2water] correctly convert index expressions for batched mma (#1168)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-25T08:54:04Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "f5f330db3d50a944048e6b75b65c8331d0ce4712",
"body": "Pywave may insert broadcasts and reshapes during the index inference and\nthey won't have water_id, or index expressions for that matter.",
"is_bot": false,
"headline": "[wave2water] handle missing water_id gracefully (#1166)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-25T08:21:02Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "bef3b82c5502504f9d6cc1ef5009562ab36d4736",
"body": "Set the co-called thread-independent constraints for all operations, not\nonly for privileged operations such as mmas and writes, during index\nexpression propagation. These include quantities derived from vector\nshapes, now available, or default indexing with 0 offset for dimensions\nnot involved in privileged operations and otherwise not deducible.\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>\n\n---------\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>",
"is_bot": false,
"headline": "[water] default-initialize index expressions for all ops (#1162)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-24T18:42:35Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "f223c86396b699a65b7e62e49f27fed65cf03987",
"body": "… (#1147)\n\nThe static file hardcoded ROCm 6.4, which breaks on other ROCm versions.\nUse gen-pytorch-rocm-requirements.py everywhere (IE in documentation,\nCI) instead, which auto-detects the installed ROCm version and probes\nfor a matching PyTorch wheel source.\n\nThis is a follow-up to the discussion in\nhttps://github.com/iree-org/wave/pull/1134\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "remove the static pytorch-rocm-requirements.txt, always use generator…",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-03-24T17:23:46Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "583f4b11c799859ec88fa0d8e34f02c8e8f4a720",
"body": "Fixes the xfail tests for streamk.\n\nCommit 3865d264 introduced unconditional linearization of global-memory\nwrites, but the linearized offsets are computed from symbolic strides\n(strides_from_symbolic_shape) which can differ from the physical layout\nstrides. Memories with an explicit MemoryLayout --\n[…]\n linearization when the\nmemory's physical_layout shape differs from its symbolic shape, falling\nback to correct multi-dimensional stores.\n\n---------\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "[compiler] Guard write linearization (#1095)",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-03-24T17:22:50Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "6765b19b90ba17b06505bf67e21f7a8322a435ac",
"body": "PackOp is a register allocation directive: its N inputs must form a\ncontiguous register block matching the pack result. Previously, pack\ninputs got independent allocations to arbitrary registers while the\nresult got a correct contiguous allocation, leaving downstream consumers\nreading uninitialized \n[…]\nr properly handling SRD construction instead of\nusing raw asm ops.\n\n---------\n\nSigned-off-by: Ivan Butygin <ivan.butygin@gmail.com>\nCo-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>",
"is_bot": false,
"headline": "[asm] Handle PackOp inputs in regalloc liveness and linear scan (#1144)",
"author_name": "Ivan Butygin",
"author_login": "Hardcode84",
"committed_at": "2026-03-24T17:14:55Z",
"body_truncated": true,
"is_coding_agent": true
},
{
"oid": "2320d7daef77c05b762efb0159ffd451031bfd1e",
"body": "Add propagation of index expressions across reductions, which is a\nprerequirement for inferring index expression in attention kernels. This\ndoesn't yet add initialization for index expressions, which isn't needed\nfor this use case.",
"is_bot": false,
"headline": "[water] propagate index expressions across reductions (#1161)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-24T15:56:21Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "231d6cc1252dc2ec4456ccc3a42b4d80eba30156",
"body": "…unrolled loops (#1175)\n\nFix G2S guard formula for eliminate_epilogue with unrolled loops\nThe guard_g2s_with_bounds_check function used (num_stages-1)*step as\nthe prefetch offset, which is incorrect when loop unrolling is active.\nWith step=2 (unroll factor 2) and a 3-stage pipeline, this produced\nIV\n[…]\n -- all with ee + sched + bufops\n```\n\n---------\n\nSigned-off-by: Sanket Pandit <sanket.pandit@amd.com>\nSigned-off-by: xintin <gaurav.verma@amd.com>\nCo-authored-by: Sanket Pandit <sanket.pandit@amd.com>",
"is_bot": false,
"headline": "[Cherry-pick][asm] Fix G2S guard formula for eliminate_epilogue with …",
"author_name": "Gaurav Verma",
"author_login": "xintin",
"committed_at": "2026-03-24T15:34:31Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "0e153e3fcd9ffe00739f6599ea68e66df2c118a5",
"body": "Diff:\nhttps://github.com/iree-org/iree/compare/iree-3.11.0...iree-3.12.0rc20260324\n\nAuto-generated by GitHub Actions using\n[`.github/workflows/update_iree_requirement_pins.yml`](https://github.com/iree-org/wave/blob/main/.github/workflows/update_iree_requirement_pins.yml).\n\nSigned-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>\nCo-authored-by: iree-pr-automator[bot] <iree-github-actions-bot@google.com>",
"is_bot": true,
"headline": "Bump IREE requirement pins to their latest versions. (#1177)",
"author_name": "iree-pr-automator[bot]",
"author_login": "iree-pr-automator[bot]",
"committed_at": "2026-03-24T14:57:43Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "b5be470f11db64042e930f865ff6b1f218e875a0",
"body": "These two quantities are related: vector shapes indicate the (desired)\nshape of the data processed by each operation instance after\ndistribution and unrolling, index expressions the position of the that\nin the larger tensor. In some cases, they may influence each other\nduring propagation. In future, we should consider folding them into one\ndata structure but for now be consistent with pywave.",
"is_bot": false,
"headline": "[water] propagate vector shapes along index expressions (#1160)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-24T14:42:57Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "185c538a57013c29b3a918e9d864bb38637284ec",
"body": "Simplifies buffer GEP handling in `translateFromLLVM` and properly folds\nbare-pointer offsets into SRD base addresses.\n\n- **Unify buffer GEP paths**: Instead of propagating bare-pointer\noffsets through `setBaseOffset` into a special-case branch in\n`handleGEP`, collapse the two buffer GEP code paths \n[…]\ntlane_b32` (or a pair for i64);\nno-op passthrough for SGPR inputs.\n\n---------\n\nSigned-off-by: Ivan Butygin <ivan.butygin@gmail.com>\nCo-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>",
"is_bot": false,
"headline": "[asm] Improve TranslateFromLLVMDialect GEP handling (#1173)",
"author_name": "Ivan Butygin",
"author_login": "Hardcode84",
"committed_at": "2026-03-24T14:41:53Z",
"body_truncated": true,
"is_coding_agent": true
},
{
"oid": "20dd2b2cf86134f996c3da093fb39220b8ed61dc",
"body": "Default linking with BFD takes unnecessarily long, bias towards LLD.",
"is_bot": false,
"headline": "[water] improve build description (#1176)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-24T12:38:02Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "06e1116b8d9b2dcbbd1e599ce6f9d2af38ced91c",
"body": "MMA ops require specific index expressions set during initialization of\nthe analysis. We shouldn't allow these values to be overwritten, even\nwith indexes defined by other MMA ops to avoid indexing and hit the\nright instruction. If the index for an MMA operand specified at its\ndefinition point diffe\n[…]\nions rather than immediately detecting conflicts. A separate\nanalysis would then choose between the possible expressions based on\nsome criteria, materialize meaningful reshapes and drop the noop ones.",
"is_bot": false,
"headline": "[water] don't propagate index expressions through MMA ops (#1149)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-24T12:36:00Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "604f2595631ee57e9cae8362ecba20c2f78e2350",
"body": "This has been failing intermittently with no attempt to fix for multiple\nmonths. Disabling.",
"is_bot": false,
"headline": "disable flaky test (#1164)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-24T12:35:51Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "14507461df8aff493f57fb42b70b8024fcb7e936",
"body": "Existing logic would override MMA vector_shape with the user-specified\nvalue from hardware constraints. This is incorrect unless the\nuser-specified value matches the vector shape expected by the MMA\ninstruction.",
"is_bot": false,
"headline": "don't override MMA vector_shape (#1141)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-24T10:29:31Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "6fa1054e79371663aa256c101669bfb743740b82",
"body": "We don't want to accidentally break these. Fix the pre-existing\nfailures.",
"is_bot": false,
"headline": "[water] test documentation builds on CI (#1172)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-23T13:54:08Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "8e1e0e3b4e02806bde6ea697300de441ccd38806",
"body": "Several places in the index expression analysis have been directly\noverriding pre-existing lattice values, in particular for:\n\n - block arguments during forward initialization, with bottom;\n - terminator operands during backward initialization, with bottom;\n - op operands during backward initializat\n[…]\nto that is to meld initalization and visitation functions, but\nit comes with higher code complexity of visitation functions and\npotentially higher cost.\n\nSigned-off-by: Alex Zinenko <git@ozinenko.com>",
"is_bot": false,
"headline": "[water] dont blindly override index lattice values (#1148)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-23T08:43:13Z",
"body_truncated": true,
"is_coding_agent": false
},
{
"oid": "226700f0ed00fb080e5fd015cd83e87fe734405b",
"body": "They were hurting my aesthetic feelings for way too long.\n\n---------\n\nSigned-off-by: Ivan Butygin <ivan.butygin@gmail.com>\nCo-authored-by: Claude Opus 4.6 <noreply@anthropic.com>",
"is_bot": false,
"headline": "Sort lists in cmake files in pre-commit (#1084)",
"author_name": "Ivan Butygin",
"author_login": "Hardcode84",
"committed_at": "2026-03-21T00:32:39Z",
"body_truncated": false,
"is_coding_agent": true
},
{
"oid": "65601846fb4d2af78209e99ad4cd703cdb2b54c0",
"body": "Kernel execution is async, it can outlive the test function if we don't\ncheck the result or explicitly synchronize.\n\n---------\n\nSigned-off-by: Ivan Butygin <ivan.butygin@gmail.com>",
"is_bot": false,
"headline": "Disable flaky cache test (#1151)",
"author_name": "Ivan Butygin",
"author_login": "Hardcode84",
"committed_at": "2026-03-20T18:57:50Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "80e64d3c23f7980096af657ffdcb1dbe25fe6f4a",
"body": "Diff:\nhttps://github.com/iree-org/iree/compare/iree-3.11.0rc20260318...iree-3.11.0\n\nAuto-generated by GitHub Actions using\n[`.github/workflows/update_iree_requirement_pins.yml`](https://github.com/iree-org/wave/blob/main/.github/workflows/update_iree_requirement_pins.yml).\n\nSigned-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>\nCo-authored-by: iree-pr-automator[bot] <iree-github-actions-bot@google.com>",
"is_bot": true,
"headline": "Bump IREE requirement pins to their latest versions. (#1158)",
"author_name": "iree-pr-automator[bot]",
"author_login": "iree-pr-automator[bot]",
"committed_at": "2026-03-20T14:12:55Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "d23224074f2766b3c69d00e2b57f326c35ff70c0",
"body": null,
"is_bot": false,
"headline": "bump LLVM version (#1157)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-20T14:12:13Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "9f8ace936e9f442ae73530237b7073fea9e568ae",
"body": "Re: https://github.com/iree-org/wave/pull/1150\n\nSigned-off-by: William G Hatch <william@hatch.uno>",
"is_bot": false,
"headline": "add upper bound to gen-pytorch-rocm-requirements.py (#1154)",
"author_name": "William G Hatch",
"author_login": "willghatch",
"committed_at": "2026-03-19T17:09:19Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "3d8c1c6b0ea2291c2271256a4ad1f0941bd6b22a",
"body": "These are much less expensive than e2e tests and may show errors without\nwaiting for e2e to complete.",
"is_bot": false,
"headline": "Run LIT tests earlier in the CI workflow (#1152)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-19T16:35:24Z",
"body_truncated": false,
"is_coding_agent": false
},
{
"oid": "336951908cda11f2e1674878d046bfbc1b7290fa",
"body": "Reverts iree-org/wave#1134. Let's see if this somehow broke wave runtime\ntests.",
"is_bot": false,
"headline": "Revert \"fix torch specs in requirements files\" (#1150)",
"author_name": "Oleksandr \"Alex\" Zinenko",
"author_login": "ftynse",
"committed_at": "2026-03-19T16:00:46Z",
"body_truncated": false,
"is_coding_agent": false
}
],
"releases_count": 9,
"commits_last_year": 882,
"latest_release_at": "2025-12-23T17:45:20Z",
"latest_release_tag": "v3.9.1",
"releases_from_tags": false,
"days_since_last_push": 26,
"active_weeks_last_year": 40,
"days_since_latest_release": 214,
"mean_days_between_releases": 19.7
},
"community": {
"has_readme": true,
"has_license": true,
"has_description": true,
"has_contributing": true,
"health_percentage": 50,
"has_issue_template": false,
"has_code_of_conduct": false,
"has_pull_request_template": false
},
"ecosystem": {
"packages": [
{
"name": "wave-lang",
"exists": true,
"license": "Apache-2.0",
"keywords": [
"Development Status :: 5 - Production/Stable",
"License :: OSI Approved :: Apache Software License",
"Programming Language :: Python :: 3"
],
"ecosystem": "pypi",
"matches_repo": true,
"registry_url": "https://pypi.org/project/wave-lang/",
"is_deprecated": false,
"latest_version": "3.9.1",
"repository_url": "https://github.com/iree-org/wave/",
"versions_count": 9,
"total_downloads": null,
"dependents_count": null,
"deprecation_note": null,
"maintainers_count": null,
"monthly_downloads": null,
"first_published_at": "2025-07-22T05:11:58.063091Z",
"latest_published_at": "2025-12-23T18:12:55.315247Z",
"latest_version_yanked": null,
"days_since_latest_publish": 214
}
]
},
"popularity": {
"forks": 32,
"stars": 58,
"watchers": 6,
"fork_history": {
"days": [
{
"date": "2025-07-16",
"count": 4
},
{
"date": "2025-07-17",
"count": 2
},
{
"date": "2025-07-21",
"count": 2
},
{
"date": "2025-07-22",
"count": 2
},
{
"date": "2025-07-24",
"count": 3
},
{
"date": "2025-07-25",
"count": 1
},
{
"date": "2025-07-28",
"count": 1
},
{
"date": "2025-07-30",
"count": 1
},
{
"date": "2025-07-31",
"count": 1
},
{
"date": "2025-08-18",
"count": 1
},
{
"date": "2025-08-20",
"count": 1
},
{
"date": "2025-09-04",
"count": 1
},
{
"date": "2025-09-09",
"count": 2
},
{
"date": "2025-09-24",
"count": 1
},
{
"date": "2025-11-12",
"count": 1
},
{
"date": "2026-01-07",
"count": 1
},
{
"date": "2026-02-07",
"count": 1
},
{
"date": "2026-02-11",
"count": 1
},
{
"date": "2026-02-21",
"count": 1
},
{
"date": "2026-03-13",
"count": 1
},
{
"date": "2026-03-30",
"count": 1
},
{
"date": "2026-04-14",
"count": 1
},
{
"date": "2026-04-25",
"count": 1
}
],
"complete": true,
"collected": 32,
"total_forks": 32
},
"star_history": null,
"open_issues_and_prs": 174
},
"ai_readiness": {
"has_nix": false,
"example_dirs": [
"examples",
"notebooks"
],
"has_llms_txt": false,
"has_dockerfile": true,
"has_mcp_signal": false,
"bootstrap_files": [],
"api_schema_files": [],
"has_devcontainer": true,
"typecheck_configs": [],
"toolchain_manifests": [
"wave_lang/kernel/wave/scheduling/aplp/Cargo.toml"
],
"largest_source_bytes": 162291,
"source_files_sampled": 528,
"oversized_source_files": 19,
"agent_instruction_files": [
".github/instructions/water-review.instructions.md"
],
"agent_instruction_max_bytes": 1071
},
"dependencies": {
"manifests": [
"pyproject.toml",
"requirements-iree-pinned.txt",
"setup.py",
"water/requirements-dev.txt"
],
"advisories": {
"error": null,
"scope": "repository_graph",
"source": "osv",
"findings": [],
"collected": true,
"malicious": [],
"truncated": false,
"by_severity": {},
"advisory_count": 0,
"affected_count": 0,
"assessed_count": 2,
"malicious_count": 0,
"assessed_package": null,
"unassessed_count": 31,
"direct_affected_count": 0
},
"ecosystems": [
"pypi"
],
"dependencies": [
{
"name": "dill",
"manifest": "pyproject.toml",
"ecosystem": "pypi",
"version_constraint": null
},
{
"name": "iree-base-compiler",
"manifest": "pyproject.toml",
"ecosystem": "pypi",
"version_constraint": "~=3.12.0rc"
},
{
"name": "iree-base-runtime",
"manifest": "pyproject.toml",
"ecosystem": "pypi",
"version_constraint": "~=3.12.0rc"
},
{
"name": "ml_dtypes",
"manifest": "pyproject.toml",
"ecosystem": "pypi",
"version_constraint": null
},
{
"name": "numpy",
"manifest": "pyproject.toml",
"ecosystem": "pypi",
"version_constraint": null
},
{
"name": "sympy",
"manifest": "pyproject.toml",
"ecosystem": "pypi",
"version_constraint": null
},
{
"name": "torch",
"manifest": "pyproject.toml",
"ecosystem": "pypi",
"version_constraint": ">=2.6,<2.10"
},
{
"name": "typing_extensions",
"manifest": "pyproject.toml",
"ecosystem": "pypi",
"version_constraint": null
}
],
"all_dependencies": {
"error": null,
"source": "github-sbom",
"packages": [
{
"name": "dill",
"direct": true,
"version": null,
"ecosystem": "pypi"
},
{
"name": "iree-base-compiler",
"direct": true,
"version": null,
"ecosystem": "pypi"
},
{
"name": "iree-base-compiler",
"direct": true,
"version": "3.12.0rc20260410",
"ecosystem": "pypi"
},
{
"name": "iree-base-runtime",
"direct": true,
"version": null,
"ecosystem": "pypi"
},
{
"name": "iree-base-runtime",
"direct": true,
"version": "3.12.0rc20260410",
"ecosystem": "pypi"
},
{
"name": "ml-dtypes",
"direct": true,
"version": null,
"ecosystem": "pypi"
},
{
"name": "numpy",
"direct": true,
"version": null,
"ecosystem": "pypi"
},
{
"name": "sympy",
"direct": true,
"version": null,
"ecosystem": "pypi"
},
{
"name": "torch",
"direct": true,
"version": null,
"ecosystem": "pypi"
},
{
"name": "typing-extensions",
"direct": true,
"version": null,
"ecosystem": "pypi"
},
{
"name": "pyo3",
"direct": false,
"version": null,
"ecosystem": "crates"
},
{
"name": "rayon",
"direct": false,
"version": null,
"ecosystem": "crates"
},
{
"name": "cmake",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "filecheck",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "lit",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "mypy",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "myst-parser",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "nanobind",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "ninja",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "pre-commit",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "pytest",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "pytest-timeout",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "pytest-xdist",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "ruff",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "setuptools",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "setuptools-rust",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "setuptools-scm",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "shibuya",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "sphinx",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "sphinx-autobuild",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "sphinx-rtd-theme",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "sphinxcontrib-mermaid",
"direct": false,
"version": null,
"ecosystem": "pypi"
},
{
"name": "wheel",
"direct": false,
"version": null,
"ecosystem": "pypi"
}
],
"collected": true,
"truncated": false,
"total_count": 33,
"direct_count": 10,
"indirect_count": 23
}
},
"maintainership": {
"issues": {
"open_prs": 81,
"merged_prs": 924,
"open_issues": 93,
"closed_ratio": 0.443,
"closed_issues": 74,
"closed_unmerged_prs": 130
},
"bus_factor": 4,
"bot_contributors": 3,
"top_contributors": [
{
"type": "User",
"login": "Hardcode84",
"commits": 262,
"avatar_url": "https://avatars.githubusercontent.com/u/3274859?v=4"
},
{
"type": "User",
"login": "harsh-nod",
"commits": 224,
"avatar_url": "https://avatars.githubusercontent.com/u/3144552?v=4"
},
{
"type": "User",
"login": "ftynse",
"commits": 181,
"avatar_url": "https://avatars.githubusercontent.com/u/1512299?v=4"
},
{
"type": "User",
"login": "raikonenfnu",
"commits": 108,
"avatar_url": "https://avatars.githubusercontent.com/u/68087699?v=4"
},
{
"type": "User",
"login": "tgymnich",
"commits": 83,
"avatar_url": "https://avatars.githubusercontent.com/u/7985149?v=4"
},
{
"type": "User",
"login": "willghatch",
"commits": 67,
"avatar_url": "https://avatars.githubusercontent.com/u/4922506?v=4"
},
{
"type": "User",
"login": "tyb0807",
"commits": 63,
"avatar_url": "https://avatars.githubusercontent.com/u/23738463?v=4"
},
{
"type": "User",
"login": "stellaraccident",
"commits": 60,
"avatar_url": "https://avatars.githubusercontent.com/u/494208?v=4"
},
{
"type": "User",
"login": "martin-luecke",
"commits": 51,
"avatar_url": "https://avatars.githubusercontent.com/u/16363540?v=4"
},
{
"type": "User",
"login": "xintin",
"commits": 46,
"avatar_url": "https://avatars.githubusercontent.com/u/48321602?v=4"
}
],
"contributors_sampled": 61,
"top_contributor_share": 0.173
},
"quality_signals": {
"has_ci": true,
"has_tests": true,
"ci_workflows": [
"build_and_release.yml",
"ci-gpu.yaml",
"ci-happy.yml",
"ci.yaml",
"perf.yaml",
"pre-commit.yaml",
"pypi-publish.yml",
"update_iree_requirement_pins.yml"
],
"has_docs_dir": true,
"linter_configs": [],
"has_editorconfig": false,
"has_linter_config": true,
"has_precommit_config": true
},
"security_signals": {
"lockfiles": [],
"scorecard": null,
"has_codeql_workflow": false,
"has_security_policy": false,
"has_dependabot_config": true
},
"contribution_flow": {
"collected": true,
"ci_last_run_at": "2026-07-26T07:42:23Z",
"oldest_open_prs": [
{
"number": 87,
"created_at": "2025-07-24T08:45:34Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 93,
"created_at": "2025-07-24T21:23:41Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 101,
"created_at": "2025-07-25T17:22:54Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 105,
"created_at": "2025-07-26T00:39:05Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 140,
"created_at": "2025-07-31T19:10:24Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 163,
"created_at": "2025-08-05T03:10:36Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 182,
"created_at": "2025-08-09T00:20:52Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 198,
"created_at": "2025-08-13T20:12:46Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 231,
"created_at": "2025-08-22T21:20:49Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 247,
"created_at": "2025-08-27T22:58:56Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 312,
"created_at": "2025-09-23T13:43:27Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 330,
"created_at": "2025-10-01T01:51:18Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 342,
"created_at": "2025-10-07T00:30:47Z",
"last_comment_at": "2025-10-07T22:12:43Z",
"last_comment_author": "harsh-amd"
},
{
"number": 347,
"created_at": "2025-10-08T20:02:42Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 451,
"created_at": "2025-11-15T01:34:01Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 490,
"created_at": "2025-11-26T23:22:29Z",
"last_comment_at": "2025-12-01T12:06:08Z",
"last_comment_author": "ftynse"
},
{
"number": 500,
"created_at": "2025-11-30T19:35:25Z",
"last_comment_at": "2025-12-01T12:03:03Z",
"last_comment_author": "ftynse"
},
{
"number": 534,
"created_at": "2025-12-05T13:04:32Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 542,
"created_at": "2025-12-08T16:44:19Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 560,
"created_at": "2025-12-10T16:51:21Z",
"last_comment_at": "2025-12-11T00:22:40Z",
"last_comment_author": "ScottTodd"
}
],
"last_merged_pr_at": "2026-04-27T20:50:55Z",
"ci_last_conclusion": "SUCCESS",
"oldest_open_issues": [
{
"number": 63,
"created_at": "2025-07-22T07:23:47Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 67,
"created_at": "2025-07-22T10:15:17Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 92,
"created_at": "2025-07-24T14:34:12Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 97,
"created_at": "2025-07-25T02:11:26Z",
"last_comment_at": "2025-08-21T07:54:29Z",
"last_comment_author": "yichiche"
},
{
"number": 106,
"created_at": "2025-07-26T01:16:04Z",
"last_comment_at": "2025-09-03T01:24:15Z",
"last_comment_author": "yichiche"
},
{
"number": 142,
"created_at": "2025-07-31T21:30:14Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 143,
"created_at": "2025-08-01T01:33:48Z",
"last_comment_at": "2025-08-20T23:31:20Z",
"last_comment_author": "paulzzy"
},
{
"number": 171,
"created_at": "2025-08-07T12:26:17Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 172,
"created_at": "2025-08-07T12:30:28Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 173,
"created_at": "2025-08-07T12:37:54Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 189,
"created_at": "2025-08-11T18:24:44Z",
"last_comment_at": "2025-08-16T17:55:17Z",
"last_comment_author": "tyb0807"
},
{
"number": 199,
"created_at": "2025-08-13T23:52:09Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 208,
"created_at": "2025-08-18T23:01:00Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 211,
"created_at": "2025-08-19T09:27:48Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 230,
"created_at": "2025-08-22T16:32:42Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 304,
"created_at": "2025-09-16T18:20:32Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 314,
"created_at": "2025-09-23T15:28:16Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 337,
"created_at": "2025-10-03T11:59:19Z",
"last_comment_at": "2025-10-03T18:57:20Z",
"last_comment_author": "Hardcode84"
},
{
"number": 458,
"created_at": "2025-11-18T16:49:14Z",
"last_comment_at": null,
"last_comment_author": null
},
{
"number": 459,
"created_at": "2025-11-18T23:18:29Z",
"last_comment_at": null,
"last_comment_author": null
}
]
}
},
"config": {
"disabled_metrics": [],
"disabled_categories": [],
"disabled_components": {}
},
"source": {
"url": "https://github.com/iree-org/wave",
"host": "github.com",
"name": "wave",
"owner": "iree-org"
},
"metrics": {
"overall": {
"key": "overall",
"band": "good",
"name": "Overall health",
"note": null,
"notes": [],
"value": 70,
"inputs": {
"security": 46,
"vitality": 81,
"community": 57,
"governance": 76,
"engineering": 81
},
"components": []
},
"categories": [
{
"key": "vitality",
"band": "good",
"name": "Vitality",
"value": 81,
"weight": 0.22,
"metrics": [
{
"key": "development_activity",
"band": "good",
"name": "Development activity",
"note": null,
"notes": [],
"value": 83,
"inputs": {
"commits_last_year": 882,
"human_commit_share": 0.89,
"days_since_last_push": 26,
"active_weeks_last_year": 40
},
"components": [
{
"key": "push_recency",
"name": "Push recency",
"detail": "last push 26 days ago",
"points": 28.8,
"status": "partial",
"details": [
{
"code": "push_recency",
"params": {
"days": 26
}
}
],
"max_points": 36
},
{
"key": "commit_cadence",
"name": "Commit cadence",
"detail": "40/52 weeks with commits",
"points": 27.7,
"status": "partial",
"details": [
{
"code": "commit_cadence_weeks",
"params": {
"weeks": 40
}
}
],
"max_points": 36
},
{
"key": "commit_volume",
"name": "Commit volume",
"detail": "882 commits in the last year",
"points": 18,
"status": "met",
"details": [
{
"code": "commits_last_year",
"params": {
"count": 882
}
}
],
"max_points": 18
},
{
"key": "openssf_scorecard_maintained",
"name": "OpenSSF Scorecard: Maintained",
"detail": "OpenSSF Scorecard unavailable",
"points": 0,
"status": "excluded",
"details": [
{
"code": "no_data",
"params": {}
}
],
"max_points": 10
}
]
},
{
"key": "release_discipline",
"band": "good",
"name": "Release discipline",
"note": null,
"notes": [],
"value": 78,
"inputs": {
"releases_count": 9,
"latest_release_tag": "v3.9.1",
"releases_from_tags": false,
"days_since_latest_release": 214,
"mean_days_between_releases": 19.7
},
"components": [
{
"key": "ships_releases",
"name": "Ships releases",
"detail": "9 releases published",
"points": 27,
"status": "met",
"details": [
{
"code": "releases_published",
"params": {
"count": 9
}
}
],
"max_points": 27
},
{
"key": "release_recency",
"name": "Release recency",
"detail": "latest release 214 days ago",
"points": 16.2,
"status": "partial",
"details": [
{
"code": "release_recency",
"params": {
"days": 214
}
}
],
"max_points": 36
},
{
"key": "release_cadence",
"name": "Release cadence",
"detail": "a release every ~19.7 days",
"points": 27,
"status": "met",
"details": [
{
"code": "release_cadence",
"params": {
"gap": 19.7
}
}
],
"max_points": 27
},
{
"key": "openssf_scorecard_signed_releases",
"name": "OpenSSF Scorecard: Signed-Releases",
"detail": "OpenSSF Scorecard unavailable",
"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": "maintained",
"guards": [],
"signals": [],
"red_flag": false,
"multiplier_pct": 100,
"declared_reason": null,
"unverified_reason": null,
"unanswered_open_prs": null,
"unanswered_open_issues": null,
"days_since_last_merged_pr": null,
"days_since_last_human_commit": 96,
"days_since_last_human_commit_is_floor": false
},
"components": [
{
"key": "project_is_still_maintained",
"name": "Project is still maintained",
"detail": "last human commit 96 days ago",
"points": 100,
"status": "met",
"details": [
{
"code": "abandonment_maintained",
"params": {
"days": 96
}
}
],
"max_points": 100
}
]
}
],
"description": "Is the project alive — is code being written and are releases shipping?"
},
{
"key": "community",
"band": "moderate",
"name": "Community & Adoption",
"value": 57,
"weight": 0.18,
"metrics": [
{
"key": "popularity",
"band": "at_risk",
"name": "Popularity & adoption",
"note": null,
"notes": [],
"value": 45,
"inputs": {
"forks": 32,
"stars": 58,
"watchers": 6,
"growth_state": "unverified",
"growth_factor_pct": 100,
"growth_unverified_reason": "no_history"
},
"components": [
{
"key": "stars",
"name": "Stars",
"detail": "58 stars",
"points": 28.5,
"status": "partial",
"details": [
{
"code": "stars",
"params": {
"count": 58
}
}
],
"max_points": 60
},
{
"key": "forks",
"name": "Forks",
"detail": "32 forks",
"points": 12.4,
"status": "partial",
"details": [
{
"code": "forks",
"params": {
"count": 32
}
}
],
"max_points": 25
},
{
"key": "watchers",
"name": "Watchers",
"detail": "6 watchers",
"points": 3.9,
"status": "partial",
"details": [
{
"code": "watchers",
"params": {
"count": 6
}
}
],
"max_points": 15
}
]
},
{
"key": "community_health",
"band": "good",
"name": "Community health",
"note": null,
"notes": [],
"value": 70,
"inputs": {
"has_readme": true,
"has_license": true,
"has_contributing": true,
"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 (Apache-2.0)",
"points": 22.5,
"status": "met",
"details": [
{
"code": "license_standard",
"params": {}
},
{
"code": "license_spdx",
"params": {
"spdx": "Apache-2.0"
}
}
],
"max_points": 22.5
},
{
"key": "contributing_guide",
"name": "CONTRIBUTING guide",
"detail": null,
"points": 18,
"status": "met",
"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
}
]
}
],
"description": "Does the project have users, downloads, attention, and a welcoming setup for contributors?"
},
{
"key": "governance",
"band": "good",
"name": "Sustainability & Governance",
"value": 76,
"weight": 0.24,
"metrics": [
{
"key": "maintainer_resilience",
"band": "good",
"name": "Maintainer resilience (bus factor)",
"note": null,
"notes": [],
"value": 84,
"inputs": {
"bus_factor": 4,
"contributors_sampled": 61,
"top_contributor_share": 0.173
},
"components": [
{
"key": "bus_factor",
"name": "Bus factor",
"detail": "4 contributor(s) cover half of all commits",
"points": 43.2,
"status": "partial",
"details": [
{
"code": "bus_factor",
"params": {
"count": 4
}
}
],
"max_points": 54
},
{
"key": "commit_distribution",
"name": "Commit distribution",
"detail": "top contributor authored 17% of commits",
"points": 18.6,
"status": "partial",
"details": [
{
"code": "top_contributor_share",
"params": {
"share": 17
}
}
],
"max_points": 22.5
},
{
"key": "contributor_breadth",
"name": "Contributor breadth",
"detail": "61 contributors",
"points": 13.5,
"status": "met",
"details": [
{
"code": "contributors_sampled",
"params": {
"count": 61
}
}
],
"max_points": 13.5
},
{
"key": "openssf_scorecard_contributors",
"name": "OpenSSF Scorecard: Contributors",
"detail": "OpenSSF Scorecard unavailable",
"points": 0,
"status": "excluded",
"details": [
{
"code": "no_data",
"params": {}
}
],
"max_points": 10
}
]
},
{
"key": "responsiveness",
"band": "moderate",
"name": "Issue & PR responsiveness",
"note": null,
"notes": [],
"value": 64,
"inputs": {
"merged_prs": 924,
"open_issues": 93,
"closed_issues": 74,
"issue_closed_ratio": 0.443,
"closed_unmerged_prs": 130
},
"components": [
{
"key": "issue_resolution",
"name": "Issue resolution",
"detail": "44% of issues closed",
"points": 20.7,
"status": "partial",
"details": [
{
"code": "issues_closed_share",
"params": {
"share": 44
}
}
],
"max_points": 46.75
},
{
"key": "pr_acceptance",
"name": "PR acceptance",
"detail": "924/1054 decided PRs merged",
"points": 33.5,
"status": "partial",
"details": [
{
"code": "decided_prs_merged",
"params": {
"merged": 924,
"decided": 1054
}
}
],
"max_points": 38.25
},
{
"key": "openssf_scorecard_code_review",
"name": "OpenSSF Scorecard: Code-Review",
"detail": "OpenSSF Scorecard unavailable",
"points": 0,
"status": "excluded",
"details": [
{
"code": "no_data",
"params": {}
}
],
"max_points": 15
}
]
},
{
"key": "stewardship",
"band": "moderate",
"name": "Ownership & stewardship",
"note": null,
"notes": [],
"value": 66,
"inputs": {
"followers": 205,
"owner_type": "Organization",
"is_verified": null,
"owner_login": "iree-org",
"public_repos": 33,
"account_age_days": 1495
},
"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": "205 followers of iree-org",
"points": 16.6,
"status": "partial",
"details": [
{
"code": "owner_followers",
"params": {
"count": 205,
"login": "iree-org"
}
}
],
"max_points": 25
},
{
"key": "track_record",
"name": "Track record",
"detail": "33 public repos, account ~4 yr old",
"points": 19.3,
"status": "partial",
"details": [
{
"code": "public_repos",
"params": {
"count": 33
}
},
{
"code": "account_age_years",
"params": {
"years": 4
}
}
],
"max_points": 25
}
]
},
{
"key": "package_maintenance",
"band": "excellent",
"name": "Package maintenance",
"note": null,
"notes": [],
"value": 91,
"inputs": {
"packages": [
"wave-lang"
],
"ecosystems": "pypi",
"any_deprecated": false,
"min_days_since_publish": 214
},
"components": [
{
"key": "published_resolvable",
"name": "Published & resolvable",
"detail": "1 package(s) on pypi",
"points": 25,
"status": "met",
"details": [
{
"code": "packages_published",
"params": {
"count": 1,
"ecosystems": "pypi"
}
}
],
"max_points": 25
},
{
"key": "publish_recency",
"name": "Publish recency",
"detail": "latest publish 214 days ago",
"points": 26,
"status": "partial",
"details": [
{
"code": "publish_recency",
"params": {
"days": 214
}
}
],
"max_points": 35
},
{
"key": "version_history",
"name": "Version history",
"detail": "9 published versions",
"points": 20,
"status": "met",
"details": [
{
"code": "published_versions",
"params": {
"count": 9
}
}
],
"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": 81,
"weight": 0.2,
"metrics": [
{
"key": "engineering_practices",
"band": "excellent",
"name": "Engineering practices",
"note": null,
"notes": [],
"value": 92,
"inputs": {
"has_ci": true,
"has_tests": true,
"has_editorconfig": false,
"has_linter_config": true,
"has_precommit_config": true
},
"components": [
{
"key": "ci_workflows",
"name": "CI workflows",
"detail": "8 workflow(s)",
"points": 24,
"status": "met",
"details": [
{
"code": "ci_workflows",
"params": {
"count": 8
}
}
],
"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": 16,
"status": "met",
"details": [],
"max_points": 16
},
{
"key": "pre_commit_hooks",
"name": "Pre-commit hooks",
"detail": null,
"points": 9.6,
"status": "met",
"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": "OpenSSF Scorecard unavailable",
"points": 0,
"status": "excluded",
"details": [
{
"code": "no_data",
"params": {}
}
],
"max_points": 20
}
]
},
{
"key": "documentation",
"band": "moderate",
"name": "Documentation",
"note": null,
"notes": [],
"value": 65,
"inputs": {
"topics": [],
"has_wiki": false,
"homepage": null,
"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": null,
"points": 0,
"status": "missed",
"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": null,
"points": 0,
"status": "missed",
"details": [],
"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": 46,
"weight": 0.16,
"metrics": [
{
"key": "security_posture",
"band": "at_risk",
"name": "Security posture",
"note": "Excluded from scoring (no data or not applicable): Dependency lockfiles. Remaining weights renormalized.",
"notes": [
{
"code": "excluded_no_data",
"params": {
"components": [
"dependency_lockfiles"
]
}
},
{
"code": "weights_renormalized",
"params": {}
}
],
"value": 33,
"inputs": {
"source": "file_signals",
"lockfiles": [],
"manifests": [
"pyproject.toml",
"requirements-iree-pinned.txt",
"setup.py",
"water/requirements-dev.txt"
],
"has_codeql_workflow": false,
"has_security_policy": false,
"has_dependabot_config": true
},
"components": [
{
"key": "security_policy_security_md",
"name": "Security policy (SECURITY.md)",
"detail": null,
"points": 0,
"status": "missed",
"details": [],
"max_points": 30
},
{
"key": "dependabot_config",
"name": "Dependabot config",
"detail": null,
"points": 25,
"status": "met",
"details": [],
"max_points": 25
},
{
"key": "dependency_lockfiles",
"name": "Dependency lockfiles",
"detail": "published library — lockfiles are an application concern, not expected",
"points": 0,
"status": "excluded",
"details": [
{
"code": "lockfiles_not_expected",
"params": {}
}
],
"max_points": 25
},
{
"key": "codeql_workflow",
"name": "CodeQL workflow",
"detail": null,
"points": 0,
"status": "missed",
"details": [],
"max_points": 20
}
]
},
{
"key": "dependency_advisories",
"band": "excellent",
"name": "Dependency advisories",
"note": "Excluded from scoring (no data or not applicable): Indirect dependencies free of known advisories, No advisories left outstanding. Remaining weights renormalized. Matched 2 resolved dependencies against OSV; 31 could not be assessed (no resolved version, an unsupported ecosystem, or beyond the reported package list). This repository publishes no package the index resolves, so the repository dependency graph was assessed instead. That graph mixes development and test pins with shipped dependencies, so only the declared runtime dependencies are scored; transitive findings are reported as context and excluded from the score. Reachability is not analyzed.",
"notes": [
{
"code": "excluded_no_data",
"params": {
"components": [
"indirect_dependencies_free_of_known_advisories",
"no_advisories_left_outstanding"
]
}
},
{
"code": "weights_renormalized",
"params": {}
},
{
"code": "advisories_scope_repository",
"params": {
"assessed": 2
}
},
{
"code": "advisories_unassessed",
"params": {
"count": 31
}
},
{
"code": "advisories_repo_graph_caveat",
"params": {}
},
{
"code": "advisories_reachability",
"params": {}
}
],
"value": 100,
"inputs": {
"source": "osv",
"advisories": 0,
"affected_packages": 0,
"assessed_packages": 2,
"unassessed_packages": 31,
"affected_by_severity": "none",
"direct_affected_packages": 0
},
"components": [
{
"key": "direct_dependencies_free_of_known_advisories",
"name": "Direct dependencies free of known advisories",
"detail": "no direct dependency carries a known advisory",
"points": 35,
"status": "met",
"details": [
{
"code": "no_direct_advisories",
"params": {}
}
],
"max_points": 35
},
{
"key": "indirect_dependencies_free_of_known_advisories",
"name": "Indirect dependencies free of known advisories",
"detail": "transitive set not separable from development and test dependencies in this scope",
"points": 0,
"status": "excluded",
"details": [
{
"code": "advisories_scope_not_separable",
"params": {}
}
],
"max_points": 25
},
{
"key": "no_advisories_left_outstanding",
"name": "No advisories left outstanding",
"detail": "no advisory carries a publication date",
"points": 0,
"status": "excluded",
"details": [
{
"code": "advisories_no_publication_date",
"params": {}
}
],
"max_points": 40
}
]
},
{
"key": "malicious_dependencies",
"band": "excellent",
"name": "Malicious dependencies",
"note": null,
"notes": [],
"value": 100,
"inputs": {
"source": "osv",
"meaning": "reported as a malicious package by the OpenSSF corpus; the remedy is removal or moving off the compromised name, never an upgrade of the same artifact. Versions the registry has since pulled are listed but not scored",
"packages": [],
"red_flag": false,
"assessed_packages": 2,
"malicious_packages": 0,
"direct_malicious_packages": 0,
"withdrawn_malicious_packages": 0,
"installable_malicious_packages": 0
},
"components": [
{
"key": "no_dependency_reported_as_a_malicious_package",
"name": "No dependency reported as a malicious package",
"detail": "no dependency is reported as a malicious package",
"points": 100,
"status": "met",
"details": [
{
"code": "no_malicious_dependencies",
"params": {}
}
],
"max_points": 100
}
]
},
{
"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": 12
},
"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": "good",
"name": "AI Readiness",
"value": 72,
"weight": 0,
"metrics": [
{
"key": "ai_agent_context",
"band": "excellent",
"name": "Agent context & guidance",
"note": null,
"notes": [],
"value": 85,
"inputs": {
"has_llms_txt": false,
"legible_history_share": 0.989,
"agent_instruction_files": [
".github/instructions/water-review.instructions.md"
],
"agent_instruction_max_bytes": 1071
},
"components": [
{
"key": "agent_instructions",
"name": "Agent instructions",
"detail": ".github/instructions/water-review.instructions.md",
"points": 45,
"status": "met",
"details": [
{
"code": "file_list",
"params": {
"files": ".github/instructions/water-review.instructions.md"
}
}
],
"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": "88 of 89 human commits state their intent (structured subject or explanatory body)",
"points": 40,
"status": "met",
"details": [
{
"code": "legible_history",
"params": {
"legible": 88,
"sampled": 89
}
}
],
"max_points": 40
}
]
},
{
"key": "ai_verify_loop",
"band": "good",
"name": "Verify loop (build / test / typecheck)",
"note": null,
"notes": [],
"value": 82,
"inputs": {
"has_nix": false,
"has_tests": true,
"lockfiles": [],
"has_dockerfile": true,
"typed_language": false,
"bootstrap_files": [],
"has_devcontainer": true,
"has_linter_config": true,
"typecheck_configs": [],
"agent_commit_share": 0.15,
"toolchain_manifests": [
"wave_lang/kernel/wave/scheduling/aplp/Cargo.toml"
],
"dependency_bot_commit_share": 0.01
},
"components": [
{
"key": "one_command_bootstrap",
"name": "One-command bootstrap",
"detail": "wave_lang/kernel/wave/scheduling/aplp/Cargo.toml (toolchain convention, no task runner)",
"points": 12.6,
"status": "partial",
"details": [
{
"code": "toolchain_convention",
"params": {
"files": "wave_lang/kernel/wave/scheduling/aplp/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": 11,
"status": "met",
"details": [],
"max_points": 11
},
{
"key": "static_type_checking",
"name": "Static type checking",
"detail": null,
"points": 0,
"status": "missed",
"details": [],
"max_points": 11
},
{
"key": "reproducible_environment",
"name": "Reproducible environment",
"detail": "devcontainer, Dockerfile",
"points": 10,
"status": "met",
"details": [
{
"code": "file_list",
"params": {
"files": "devcontainer, Dockerfile"
}
}
],
"max_points": 10
},
{
"key": "demonstrated_agent_practice",
"name": "Demonstrated agent practice",
"detail": "15 of the last 100 commits agent-authored or agent-credited",
"points": 10,
"status": "met",
"details": [
{
"code": "agent_authored_commits",
"params": {
"count": 15,
"sampled": 100
}
}
],
"max_points": 10
},
{
"key": "automated_maintenance",
"name": "Automated maintenance",
"detail": "1 of the last 100 commits are automated dependency updates",
"points": 8,
"status": "met",
"details": [
{
"code": "dependency_bot_commits",
"params": {
"count": 1,
"sampled": 100
}
}
],
"max_points": 8
},
{
"key": "openssf_scorecard_pinned_dependencies",
"name": "OpenSSF Scorecard: Pinned-Dependencies",
"detail": "OpenSSF Scorecard unavailable",
"points": 0,
"status": "excluded",
"details": [
{
"code": "no_data",
"params": {}
}
],
"max_points": 10
}
]
},
{
"key": "ai_code_legibility",
"band": "moderate",
"name": "Code legibility for models",
"note": null,
"notes": [],
"value": 53,
"inputs": {
"primary_language": "Python",
"largest_source_bytes": 162291,
"source_files_sampled": 528,
"oversized_source_files": 19
},
"components": [
{
"key": "type_checkable_code",
"name": "Type-checkable code",
"detail": "Python without a type-check config",
"points": 0,
"status": "missed",
"details": [
{
"code": "no_typecheck_config_language",
"params": {
"language": "Python"
}
}
],
"max_points": 45
},
{
"key": "manageable_file_sizes",
"name": "Manageable file sizes",
"detail": "19/528 source files over 60KB",
"points": 53,
"status": "partial",
"details": [
{
"code": "oversized_source_files",
"params": {
"kb": 60,
"sampled": 528,
"oversized": 19
}
}
],
"max_points": 55
}
]
},
{
"key": "ai_interfaces",
"band": "at_risk",
"name": "Machine-readable interfaces",
"note": null,
"notes": [],
"value": 40,
"inputs": {
"example_dirs": [
"examples",
"notebooks"
],
"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, notebooks",
"points": 40,
"status": "met",
"details": [
{
"code": "file_list",
"params": {
"files": "examples, notebooks"
}
}
],
"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": [
"Star history unavailable: GitHub GraphQL error: Resource not accessible by personal access token",
"OpenSSF Scorecard did not return a usable result (exit code -9); skipping Scorecard checks"
],
"report_type": "repository",
"generated_at": "2026-07-26T15:25:15.590031Z",
"schema_version": "0.27.0",
"badge_url": "https://raw.githubusercontent.com/inspect-software/badges/main/v1/i/iree-org/wave.svg",
"full_name": "iree-org/wave",
"license_state": "standard",
"license_spdx": "Apache-2.0"
}