agentmods.dev

autoresearch

autoresearch is a skill for Claude Code from tdimino/claude-code-minoan. It costs 82 tokens per session (2,385 once invoked), scanned A, original, MIT.

A setup for repeated, automatic code-improvement experiments in a Git repository. Each experiment proposes one change, measures it with a fixed evaluation, and keeps or discards the change based on the result.

In plain words
What is it for?
Use it to scaffold improvement loops, create evaluation checks, run competing code hypotheses, and track progress across experiments.
Why use it?
It replaces subjective decisions about whether a change helped with a repeatable keep-or-discard process and a Git history of experiments.

Skill for Claude Code

Written for Claude Code: argument-hint in frontmatter. Also seen: reads .claude/ paths.

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is python3 .lab/eval.py.

not rated 41repo yesterday A scan SkillSpector: warn 82 tokens original MIT

Good fit Use it to scaffold improvement loops, create evaluation checks, run competing code hypotheses, and track progress across experiments.

Compare 6 skills from other repositories ↓
View source ↗ tdimino/claude-code-minoan minoanmystery.org
Install

Getting it into your agent

It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.

Clone the repo
git clone --depth 1 https://github.com/tdimino/claude-code-minoan
agentmods
npx agentmods add skills/tdimino/claude-code-minoan/autoresearch

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for autoresearch

README.md
[![agentmods](https://agentmods.dev/badge/skills/tdimino/claude-code-minoan/autoresearch/github.svg)](https://agentmods.dev/skills/tdimino/claude-code-minoan/autoresearch)
Your own site
<a href="https://agentmods.dev/skills/tdimino/claude-code-minoan/autoresearch"><img src="https://agentmods.dev/badge/skills/tdimino/claude-code-minoan/autoresearch/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for autoresearch

Your own site · 80×15
<a href="https://agentmods.dev/skills/tdimino/claude-code-minoan/autoresearch"><img src="https://agentmods.dev/badge/skills/tdimino/claude-code-minoan/autoresearch.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 82 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,385 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector warn 7 Sept 2026
SkillSpector: 2 findings, up to high

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • high Tool Misuse · line 19
    Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).
    Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
  • high Tool Misuse · line 182
    Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).
    Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00082 $0.02385
Opus 5 $0.00041 $0.01192
Sonnet 5 $0.00016 $0.00477
Haiku 4.5 $0.00008 $0.00238

Measured 12d ago against content hash 56280c7552e2, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

autoresearch scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.

The scan reads SKILL.md. This mod also ships 6 executable files (assets/eval_base.py, scripts/detect_stack.py, scripts/eval_gen.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/core-development/autoresearch/SKILL.md · 236 lines

How it starts

The opening of the file, as written. The whole thing — 236 lines — stays where its author put it; the contents beside it link to each section on GitHub.

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)

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

Runbooks — mechanical process with clear steps, not cognitive reasoning.

Quick Start

/autoresearch init          # scaffold .lab/ in your repo
/autoresearch run           # start the loop (default: 50 iterations)
/autoresearch status        # check progress
/autoresearch resume        # recover interrupted run

Command Dispatch

Parse $ARGUMENTS and route:

Argument Action
init Run scaffold workflow (see Init below)
eval-gen Regenerate eval gates from repo analysis
run [--max-iterations N] [--dry-run] Launch the autoresearch loop
status Show composite, timeline, convergence signals
resume Detect .lab/, present state, ask resume or fresh
(empty) Show help text with available commands

Init Workflow (/autoresearch init)

  1. Verify .git/ exists in current directory
  2. Run stack detection: 
    python3 ~/.claude/skills/autoresearch/scripts/detect_stack.py
  3. Review the detected stack info (language, build_cmd, test_cmd, lint_cmd)
  4. Run the scaffold script: 
    python3 ~/.claude/skills/autoresearch/scripts/scaffold.py --repo-root . --yes
  5. Review .lab/config.json — adjust keep_threshold, max_iterations, gate_weights if needed
  6. Edit .lab/program.md — this is the most important file. Add:
    • Specific areas to improve (not vague goals)
    • Concrete hypothesis list (ranked)
    • Constraints the agent must respect
  7. Run baseline eval to verify gates work: 
    python3 .lab/eval.py
  8. Report the initial composite to the user

Read the full file on GitHub · 236 lines

Files

What ships with it

12 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 12d ago First seen · 236 lines · 82 tokens per session scan A 56280c7552e2

Subscribe to this mod's changes

autoresearch is a skill published in the GitHub repository tdimino/claude-code-minoan (41 stars, last pushed yesterday), licensed MIT. It adds 82 tokens to every session and 2,385 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Claude CodeResearchTestingclaude-code-skillscli-appminoanopen-source
Related

Other skills, from other repositories

test-native-extension

Validate a third-party control repo across four automated layers plus one printed manual recipe. Layer 1 asserts native-source structure (Android getName() and iOS +moduleName to manifest nativeModule; @ReactMethod / RCTEXPORTMETHOD to methods; no @ReactModule) plus load/init readiness (ReactPackage public no-arg…

microsoft/power-platform-skills · 211 tokens

test-site

Tests a deployed, activated Power Pages site at runtime using browser-based navigation, page crawling, and API request verification via Playwright. Use when the user wants to test, verify, or smoke-test their deployed site.

microsoft/power-platform-skills · 46 tokens

simulation-study

Scaffold and run a reproducible Monte Carlo simulation study in R — a declared assumption regime, a parameterized DGP, an estimator grid, a seeded replication loop, and a summary of bias, RMSE, empirical SE, coverage, size/power with Monte Carlo standard errors. Use when the user says "run a Monte Carlo simulation"…

pedrohcgs/claude-code-my-workflow · 148 tokens

test-quality

Write high-quality JUnit 5 tests with AssertJ assertions. Use when user says "add tests", "write tests", "improve test coverage", or when reviewing/creating test classes for Java code.

decebals/claude-code-java · 45 tokens

voice-agent-test-harness

Drive a fixed suite of spoken tests against a voice agent ("subject") from a co-located machine ("prober"), measure response latency / clarity / accuracy, diff against baseline, and report to the owner over Telegram.

sonichi/sutando · 0 tokens

054-design-tdd

Use when Java implementation work should be guided by Test-Driven Development, including maintaining a test list, choosing the next behavior, writing a failing test first, implementing only enough production code to pass, and refactoring while keeping tests green. This should trigger for requests such as Apply TDD…

jabrena/plinth · 93 tokens