#llm

Herrscher Backend plugin: the only module that talks to Claude — implements the contracts.Backend port, streaming Claude Code stream-json turns.

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

No repository description published.

MCP server giving AI coding agents offline, token-efficient persistent memory — recall decisions, patterns, and bugs across sessions with 68.5% fewer tokens than re-reading context. No LLM, no server; works with 15+ agents.

Run Codex CLI with OpenAI-compatible providers and orchestrate profile-based agents from Python

Self-improving AI agent harness — propose, eval, approve, ship. AHE autonomous loop + MCP + BYOK + eval suite.

Agent framework and applications built upon Qwen>=3.0, featuring Function Calling, MCP, Code Interpreter, RAG, Chrome extension, etc.

P2P multi agent orchestration harness, universal automation system and IDE

Jurisdiction-aware judicial reasoning platform — a model-agnostic AI engine that produces citation-backed, adversarially-tested draft case analyses for human sign-off. Non-binding by design.

Accurate, current LLM model identifiers and capabilities for every provider — shipped as data, refreshed from official sources.

Vendor-neutral Go CLI for validating OpenAI-compatible /models and opt-in /chat/completions behavior.

beecork — a tiny CLI coding agent built from scratch (TypeScript, OpenRouter)

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.

Research driven cognitive memory framework for AI systems.

Unattended Go daemon that runs an LLM coding agent against a queue of tasks — it syncs, executes, verifies, and reports back on a loop.

TypeScript SDK for Declaw — security-first sandboxing for AI agents

EigenPal Typescript SDK

Rate limiting, caching, and request prioritization for modern workloads

Local-first semantic governance for AI agents: contracts, impact, drift, evidence, and verified execution. Codex is the first native integration.

No repository description published.

Capability-aware vision + paste extension for the pi coding agent. Delegates image analysis only when the active model is text-only; passes through natively for multimodal models (zero delegation).

Google Gemini MCP - Use Google's visual, video and chat capabilities in your AI Assistant. Uses MCP Apps to preview created images, landing pages, SVGs and video

No repository description published.

Auditable RAG from the command line: grounded, cited answers over your own documents, with faithfulness verification and retrieval eval you can gate CI on.

Persistent memory for AI coding agents — your agent stops forgetting between sessions. One Go binary, one SQLite file.

A CLI tool that converts database schemas into the Toon schema definition format.

One Agent Client Protocol surface, three CLI backends. Drive Claude Code, codex, and opencode without backend-specific orchestrator code.

Your AI finds the open-source tool it needs on GitHub. Legwork installs it in a sandbox.

No repository description published.

Guardrails for AI agents, with a receipt. Blocks destructive tool calls before they run and signs every decision into a hash-chained record anyone can verify with an open-source tool.