#llm

Vector Graph RAG for code & docs with Persistent Chat-Session Memory — BM25 (multi-lang), AST chunking. Hybrid, Compute, Scan, Timeline, Absence, Salience. Graph + temporal knowledge. Privacy Stamps. Source Adapters. Chat summarisation + verbatim copy + free scan. Git indexing. CLI, MCP and Claude Code plugin. Ollama or cloud LLMs. Dashboard.

Uncensored AI coding agent for your terminal — Claude Code alternative with MCP support. Drives IDA Pro, Roblox Studio, Wireshark, Blender, and any MCP server.

Test and evaluate LLMs and model configurations, across all the scenarios that matter for your application

Claude Code 多模型代理 — 一键切换 DeepSeek、Qwen、GLM、Moonshot 等国产大模型，协议自动转换，零配置即用

Webbee Code 🐝 — the imperal.io Cloud coding ultra-agent in your terminal

Self-hosted observability for LangGraph — traces, cost, sessions, scores, error forensics, and a living graph view. pip install windhover

Slim SDK for Soothe plugins and shared contracts (events, wire, display, protocols)

A multi-agent skill package (Skill OS) for AI coding assistants (Claude Code, Cursor, Windsurf, Antigravity) with 23 professional skills.

Privacy-first CLI that transforms Jira tickets into AI-ready briefs. 60–80% fewer tokens. No relay. Pipe-friendly.

Cross-session, traceable, self-hostable persistent memory for AI agents.

A coding-discipline spec for OpenAI's Codex CLI, enforced by native Codex hooks + a rule-hit telemetry loop. Turns an AGENTS.md the model drifts from into a system that holds. Independent of oh-my-codex.

Official Cadreen SDKs - TypeScript, Go, and Python

Rust builders for xgrammar structural_tag tool-calling constraints

Go-native, zero-dependency LLM command-output compressor — an embeddable take on rtk (the Rust 'token killer') with lossless-by-default compression and credential redaction.

Robust extraction of structured payloads from LLM output via sentinel blocks (<<<TAG>>> … <<<END>>>). Survives quotes, newlines & braces that break JSON.parse. 10 language ports, one spec.

Deterministic, evidence-based build harness for autonomous coding agents — FSM sprint gates, execution-evidence verification, and adversarial critics. Proof, not opinion.

AI 能力底座 — 统一 LLM Provider 接口（OpenAI/Anthropic/DeepSeek/Gemini/Ollama），图片生成，视频生成，Embedding，向量存储，记忆管理。Hexagon 生态 L1。

MCP server for Toss Securities (토스증권) Open API: Korean stock (KRX) quotes, holdings, orders & technical analysis for Claude and other AI agents. Read-only by default.

Supercharged vibe coding — local proxy that auto-routes Claude Code requests to any provider with transparent API translation, smart routing, and failover. Single Go binary with super ease to use TUI config wizard and live usage monitor.

Multi-provider LLM router for Go — retries, fallbacks, and streaming across OpenAI, Anthropic, Gemini, and more

A schema-first GraphQL MCP server that avoids the 1MB introspection limit via typeNames/typeDepth scoping — built for federated topologies.

A minimal, extensible agent library for Python.

m8tes SDK. Ship production-ready agents in seconds. Skip the infrastructure. Hosted runtime, computer use, 150+ integrations, scheduling, and memory. Ship agents that work like people.

Minimal, zero-dependency provider switching for Claude Code — switch between Anthropic, OpenRouter, DeepSeek, Z.AI, Kimi, Qwen, MiniMax and custom endpoints

macOS local-first chat & agent tool with interactive data analysis (Wails v2 + React)

Parse partial JSON generated by LLM

Portable, cross-agent memory & workflow for AI coding agents — Claude Code, Codex, Cursor, Devin, and more. Monorepo: memory substrate, the kit (composition root), CLI bridges, and the engine.

Make AI agents follow instructions reliably and prove what they did. State machine engine with evidence chains, multi-agent delegation, human gates, and auditable execution — via MCP or CLI.

Offline CLI for per-prompt coding-assistant token usage

Cortx — AI Agent runtime | Harness Agent + Agentic Loop | 43 tools | Python + TypeScript