Scaffold and run Karpathy-style autoresearch loops in any git repo. This skill should be used when setting up autonomous code improvement, generating adversarial eval harnesses, running hypothesis-implement-eval-keep/discard loops, or checking autoresearch progress. Triggers on "autoresearch", "autonomous improvement", "eval loop", "hypothesis loop", "self-improvement loop".
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
<critical>
## Five Invariants (never violate)
1. **Single mutable surface** — one hypothesis per iteration, one change per experiment
2. **Fixed eval budget** — eval runs in bounded time, no network calls in gates
3. **One scalar metric** — composite score drives keep/discard, not vibes
4. **Binary keep/discard** — improved = keep, else revert `git reset --hard HEAD~1`
5. **Git-as-memory** — every experiment is a commit, discards are reverts, history is the log
## Safety rules
- Never modify `.lab/` contents during hypothesis implementation
- Never skip eval — every commit must be evaluated before keep/discard
- Always revert on crash — `atexit` handler restores git state
- Runner uses **subscription auth** (`claude -p` with ANTHROPIC_API_KEY stripped)
</critical>
# Autoresearch
Scaffold and run autonomous code improvement loops in any git repo. The pattern: generate a hypothesis via `claude -p`, implement it, run programmatic eval gates, keep if the composite score improves, discard if it doesn't. Proven across 50+ iterations on two codebases (shadow-engine: 0.69 to 1.0, perplexity-clone: search quality optimization).
## Category
…
The line at the top of this page pins the label git:20260330.e8b27c0, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
GET https://markdownregistry.com/api/v1/artifacts/art_k5vtfvki7omis3ux
GET https://markdownregistry.com/api/v1/resolve?ref=tdimino/claude-code-minoan/autoresearch
GET https://markdownregistry.com/api/v1/blob/56280c7552e22b5c26b8c0c187eab00a8c805d3154378b27a3c7c0a36a10a95e
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
tdimino/claude-code-minoan · skills/core-development/agents-md-manager/SKILL.md · This skill should be used when creating, auditing, or maintaining AGENTS.md files and Codex CLI configuration for any…
tdimino/claude-code-minoan · skills/core-development/claude-md-manager/SKILL.md · Create, audit, and maintain CLAUDE.md documentation files that configure Claude Code for projects. Use this skill when…
tdimino/claude-code-minoan · skills/core-development/claude-usage/SKILL.md · Report ground-truth Claude token consumption and estimated cost by parsing JSONL session files directly. Use when…
tdimino/claude-code-minoan · skills/core-development/dag-typesafe/SKILL.md · Analyze a repository's type system and generate type-safe DAG execution pipelines with GraphSentry-style certificate…
tdimino/claude-code-minoan · skills/core-development/glossary/SKILL.md · Build or update a CONTEXT.md domain glossary for the current project. Establishes shared vocabulary with term…
tdimino/claude-code-minoan · skills/core-development/openrouter-usage/SKILL.md · Query OpenRouter API costs, credits, and usage breakdown by model, provider, and date for the last 30 days. Triggers on…
github/awesome-copilot · skills/autoresearch/SKILL.md · Autonomous iterative experimentation loop for any programming task. Guides the user through defining goals, measurable…
orchestra-research/ai-research-skills · 0-autoresearch-skill/SKILL.md · Orchestrates end-to-end autonomous AI research projects using a two-loop architecture. The inner loop runs rapid…
ericosiu/ai-marketing-skills · autoresearch/SKILL.md · Run Karpathy-style autoresearch optimization on any content. Generates 50+ variants, scores with a 5-expert simulated…
yeachan-heo/gajae-code · packages/coding-agent/src/defaults/gjc/skills/autoresearch/SKILL.md · Goal-directed research missions that interleave web research with data experimentation and end on a structured…
