LLM-as-a-Verifier is a general-purpose framework that provides fine-grained feedback for any agent without requiring additional training. It achieves SOTA performance across coding, robotics, and medical agentic benchmarks.
Repository Index
73,862 visible repositories in the index
49–72 of 118 · Most stars
Intercept and inspect Coding Agent API traffic from Claude Code, Codex CLI, Gemini CLI, Cursor CLI, OpenCode, Kimi/Kimi Code, Pi, and Hermes in a local trace viewer.
Templates and workflow for generating PRDs, Tech Designs, and MVP and more using LLMs for AI IDEs
- ai-agents
- ai-workflow
- beginner-friendly
- mvp-development
Same agent Smaller bill. The tools and infrastructure to get more work from your AI. Fewer tokens. Less compute. Lower cost.
- observability
- llmops
- ai-observability
- openai
OpenGame: Open Agentic Coding for Games
Open-source desktop AI agent for tools, files, knowledge, workflows, and real deliverables.
Lucebox inference engine for heterogeneous hardware & consumer GPUs
Find the ghost tokens. Fix them. Survive compaction. Avoid context quality decay.
- claude-code
- token-optimization
- context-window
- token-usage
Find, benchmark and install in CLI 170+ FREE coding LLM models across 15+ providers in real time
Formal verification for AI: a Lean 4 + Dafny proof checker that catches fake AI proofs (reward hacking), and verified recursive self-improvement (RSI) for coding agents. AI for math, physics and software engineering.
- proof-assistant
- lean4
- ai-agents
- ai-for-math
The fastest way to run Qwen 3.8 Flash Next, Qwen 3.8 27B and Ternary Bonsai 2 27B on a Mac: 125 tok/s in OpenCode on an M5 Max, and a 27B model on 16 GB Macs. Native MTP speculative decoding on Apple Silicon, exact at any temperature. OpenAI and Anthropic compatible local server.
- apple-silicon
- local-ai
- mlx
- mtp
the holly uncodexify instructions - letting GPT create uncodexified UI
[NeurIPS 2026] Safety failures in coding agents and frontier LLMs. Reproducible workspaces for studying when AI agents generate unsafe data during task completion.
MiniMax-M2, a model built for Max coding & agentic workflows.
Research into how agentic AI coding assistants work. Reconstructed prompt patterns, agent coordination, and security classification
- agentic-ai
- ai-research
- claude
- prompt-engineering
Antigravity Manager is a desktop application for managing the Google accounts used by Antigravity clients (including Classic, Antigravity IDE and agy CLI). View model quota in real time, switch client accounts seamlessly, and connect AI coding tools like Claude Code, Codex and OpenCode through a high-performance local API gateway.
- antigravity
- antigravity-ai
- antigravity-ide
- electron
An open-source Harness Engineering platform for coding agents—define harnesses as code, run controlled experiments, inspect evidence, and compare outcomes. Turn task evidence into actionable team and organization insights.
- agent-plugin
- claude-code
- codex
- cursor
The Self-Coding System for Your App — Alan AI SDK for Web
- voice-ai
- conversational-ai
- machine-learning
- ai-chat
A curated, battle-tested AI-coding toolkit: Claude Code plugins, subagent orchestration, quality gates, and ready-to-copy prompts, extracted from real production use.
The Fable Workflow: how Claude Fable 5 worked, distilled into skills any model can run, with the eval that keeps it honest. Think / act / prove.
Web Crawling and RAG Capabilities for AI Agents and AI Coding Assistants
A local inference engine for Apple silicon, built around the model.
- apple-silicon
- coding-agents
- llm-inference
- macos
Local-first AI token usage & cost tracker for 31 coding tools incl. Claude Code, Codex, Cursor, Gemini & DeepSeek Harness—with native apps. Never reads prompts.
- claude-code
- cli
- codex-cli
- cost-tracker
Use Claude and Antigravity with Pi, OpenCode and other coding clients. Local API bridge, usage dashboard and Mac app. Antigravity preview available in 1.74.0.