Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Autonomous experiment loop skill for Claude Code — port of pi-autoresearch
| Date | Stars |
|---|---|
| 2026-08-05 | 335 |
| 2026-08-13 | 339 |
| 2026-08-18 | 340 |
| 2026-08-19 | 341 |
| 2026-08-20 | 340 |
| 2026-08-24 | 341 |
| 2026-09-01 | 342 |
| 2026-09-04 | 343 |
| 2026-09-07 | 342 |
| 2026-09-09 | 343 |
| 2026-09-17 | 344 |
| 2026-09-23 | 345 |
| 2026-10-01 | 346 |
| 2026-10-03 | 346 |
| 2026-10-05 | 345 |
| 2026-10-08 | 346 |
| 2026-10-30 | 347 |
| 2026-11-04 | 348 |
Today
+1 stars today
This week
+2 stars this week
This month
+3 stars this month
Momentum
0.0
growth rate 0.58%/day
# autoresearch-claude-code  [](LICENSE) A measured experiment loop for **Claude Code and Codex**. Define a goal, a fixed benchmark, and files to modify. The agent tries changes, measures results, commits supported improvements, and records failed ideas within a budget. Inspired by [pi-autoresearch](https://github.com/davebcn87/pi-autoresearch). The experiment protocol is a shared skill with standard-library Python helpers; no MCP server is required. ## Install Requires **Python 3.10+, Bash, Git**, and your chosen agent client. Linux and macOS are supported; use WSL on Windows. The optional ML example has its own dependencies. ```bash gh repo clone drivelineresearch/autoresearch-claude-code cd autoresearch-claude-code ./install.sh --codex # Codex skill in ~/.agents/skills/autoresearch ./install.sh --claude # Claude skill, command, and four hooks # Or: ./install.sh --all ``` No argument defaults to Claude for compatibility. Installations use symlinks, so keep this checkout in place. Existing foreign files/directories are preserved and conflicts are reported. Claude settings are validated before changes, updated atomically, and backed up when modified. Codex installation does not edit its configuration or register Claude hooks. Claude also supports a session-local plugin load: ```bash claude --plugin-dir /absolute/path/to/autoresearch-claude-code ``` Choose plugin loading or manual Claude installation; installing both can duplicate hooks. For removal, use `./uninstall.sh --codex`, `--claude`, or `--all`. Removal only deletes symlinks owned by this checkout and corresponding exact hook entries; manually copied files and other installations remain untouched. See [installation/review notes](docs/review.md) for migration limits. ## Quick start In **Codex CLI/IDE**, invoke the installed skill: ```text $autoresearch optimize test suite runtime with at most 20 runs $autoresearch status $autoresearch report $autoresearch off ``` In the Codex app, select the skill or ask to use `autoresearch` by name. Codex [discovers symlinked skills in ~/.agents/skills](https://learn.chatgpt.com/docs/build-skills). For **Claude Code**, manual installation uses `/autoresearch`; plugin loading uses the namespaced `/autoresearch:autoresearch`, with the same goal or subcommands. The agent establishes scope, creates an experiment branch, prepares a fixed benchmark, calibrates noise, and logs a baseline. It preserves existing work; use a dedicated worktree when another task is in progress. Artifacts must be ignored in the target project too; this plugin's ignores do not transfer. For unattended Codex execution, first initialize and pause the session with the skill, review the baseline, then explicitly resume it and run: ```bash # Run from the initialized experiment worktree after reviewing/removing its pause sentinel. python3 ~/.agents/skills/autoresearch/scripts/codex_loop.py --dry-run python3 ~/.agents/skills/autoresearch/scripts/codex_loop.py \ --max-turns 10 --max-seconds 3600 --turn-timeout 300 \ --protect path/to/scorer.py ``` Replace `path/to/scorer.py` with your real scorer file; repeat `--protect` for other locked inputs. See the [Codex guide](skills/autoresearch/references/codex.md) for initialization, flags, logs, pause/cancellation, and recovery. ## What is enforced | Capability | Mechanism and limit | |---|---| | Shared state | Config/result JSONL validated for finite numbers, sequential runs, segments, status, and metadata; writes locked and atomic | | Claude continuation | Stop hook blocks turn completion while an active valid session has budget; malformed state allows stopping with a diagnostic | | Codex continuation | Bounded `codex exec` supervisor requires one result per invocation; stops on no progress, failure, altered config, branch, or protected files | | Budgets | Current-segment run/time/targ
Excerpt of 7,910 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:ca8d10c1ca44a616, llm:Description: 'Autonomous experiment loop skill for Claude Code — port of pi-autoresearch' (Python).
matched fp:ca8d10c1ca44a616, llm:Description: 'Autonomous experiment loop skill for Claude Code — port of pi-autoresearch' (Python).
matched fp:ca8d10c1ca44a616, llm:Description: 'Autonomous experiment loop skill for Claude Code — port of pi-autoresearch' (Python).