HERO = Hashing · Edge cases · Rubrics · Overbuild — the four shapes coding agents over-defend in. A paste-in contract that stops them. Works with Claude Code, Codex, Antigravity, Cursor, Copilot, W…

Open Science Desktop — local-first, model-agnostic AI research workbench for macOS, Windows & Linux. Open-source Claude Science desktop alternative built on Tauri + MCP + agent skills.

A debugging framework for agentic AI systems: diagnose failures, attribute root causes, recover with evidence, and validate fixes through reruns.

🦞 Just talk to your agent — it learns and EVOLVES 🧬.

A general framework for distilling human-created multimodal resources into reusable, executable skills that AI agents can browse, compose, and run, validated across diverse domains including web, P…

Reference code for the Meta-Harness paper.

The roadmap of long-horizon agents

Training terminal-agents

This is the repo for the paper TerminalTraj: Large-Scale Terminal Agentic Trajectory Generation from Dockerized Environments

The open-source AI workbench for scientific research

Don't trust an autoresearch paper at face value. Reviewer-side integrity forensics (self-consistency + fabrication), deterministic verdict. 61 signals: 46 integrity hack-patterns (families A–H, ver…

A benchmark for evaluating AI agents on frontier ultra long-horizon auto research tasks.

Repository for TRACE: Capability-Targeted Agentic Training

[Survey] A Comprehensive Survey of Self-Evolving AI Agents: A New Paradigm Bridging Foundation Models and Lifelong Agentic Systems

Code for the multi-agent computer use project.

Open-World Self-Evolution for LLM Agents — agents that build both their skills and their own verification signals from scratch, with no target-task supervision. (Code coming soon.)

One dashboard. An entire research team.

🔥 A Survey on AI Auto-Research

🤗 ml-intern: an open-source ML engineer that reads papers, trains models, and ships ML models

🙌 OpenHands: AI-Driven Development

AIDE: an LLM agent for machine learning engineering - the research Weco grew out of. Referenced in OpenAI MLE-bench.

Dependency-Aware Structural Retrieval for Massive Agent Skills

AIRS-Bench: an AI Research Science benchmark for quantifying the end-to-end AI research abilities of LLM agents

Open-source autoresearch powered by autonomous coding agents. Run Claude Code, OpenCode, and Codex with grading, shared knowledge, and multi-agent evolution. Accepted at COLM 2026.

Now, Stronger AI Pushes Frontiers, Stronger Our Shared Future.

Dr. Claw plugin for Claude Code — AI research pipeline