autoresearch

autoresearch is a skill for Claude Code, Codex from a-tokyo/agent-skills. It costs 116 tokens per session (2,571 once invoked), scanned A, a copy of autoresearch, MIT.

An automated cycle for improving a programming task through repeated code changes, tests or experiments, measurements, and selective retention of results. It requires a defined goal and measurable outcome.

In plain words
What is it for?
Use it for programming tasks where progress can be measured, such as improving a test result or another defined metric.
Why use it?
It reduces the need to run and compare each experiment manually while preserving a baseline and reverting changes that do not improve the result.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it for programming tasks where progress can be measured, such as improving a test result or another defined metric.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/a-tokyo/agent-skills/autoresearch
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add a-tokyo/agent-skills --skill autoresearch
Clone the repo
git clone --depth 1 https://github.com/a-tokyo/agent-skills

Made for: Claude Code, Codex.

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/a-tokyo/agent-skills/autoresearch/github.svg)](https://agentmods.dev/skills/a-tokyo/agent-skills/autoresearch)
Your own site
<a href="https://agentmods.dev/skills/a-tokyo/agent-skills/autoresearch"><img src="https://agentmods.dev/badge/skills/a-tokyo/agent-skills/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/a-tokyo/agent-skills/autoresearch"><img src="https://agentmods.dev/badge/skills/a-tokyo/agent-skills/autoresearch.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 116 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,571 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.
Origin 100% copy Near-identical to another mod 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 $0.00116 $0.02571
Opus 5 $0.00058 $0.01286
Sonnet 5 $0.00023 $0.00514
Haiku 4.5 $0.00012 $0.00257

Measured 11d ago against content hash 754179e98751, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, 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 11d ago.

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.

Origin

This is a copy

100% identical to autoresearch — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

benchmarks/create-skill-autoresearch/companions/autoresearch/SKILL.md · 276 lines

How it starts

The opening of the file, as written. The whole thing — 276 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Autoresearch: Autonomous Iterative Experimentation

An autonomous experimentation loop for any programming task. You define the goal and how to measure it; the agent iterates autonomously -- modifying code, running experiments, measuring results, and keeping or discarding changes -- until interrupted.

This skill is inspired by Karpathy's autoresearch, generalized from ML training to any programming task with a measurable outcome.


Agent Behavior Rules

  1. DO guide the user through the Setup phase interactively before starting the loop.
  2. DO establish a baseline measurement before making any changes.
  3. DO commit every experiment attempt before running it (so it can be reverted cleanly).
  4. DO keep a results log (TSV) tracking every experiment.
  5. DO revert changes that do not improve the metric (git reset to last known good).
  6. DO run autonomously once the loop starts -- never pause to ask "should I continue?".
  7. DO NOT modify files the user marked as out-of-scope.
  8. DO NOT skip the measurement step -- every experiment must be measured.
  9. DO NOT keep changes that regress the metric unless the user explicitly allowed trade-offs.
  10. DO NOT install new dependencies or make environment changes unless the user approved it.

Phase 1: Setup (Interactive)

Before any experimentation begins, work with the user to establish these parameters. Ask the user directly for each item. Do not assume or skip any.

1.1 Define the Goal

Ask the user:

What are you trying to improve or optimize?

Examples: execution time, memory usage, binary size, test pass rate, code coverage, API response latency, throughput, error rate, benchmark score, build time, bundle size, lines of code, cyclomatic complexity, etc.

Record the user's answer as the goal.

1.2 Define the Metric

Ask the user:

How do we measure success? What exact command produces the metric?

I need:

  1. The command to run (e.g., dotnet test, npm run benchmark, time ./build.sh, pytest --tb=short)
  2. How to extract the metric from the output (e.g., a regex pattern, a specific line, a JSON field)
  3. Direction: Is lower better or higher better?

Example: "Run dotnet test --logger trx, count passing tests. Higher is better." Example: "Run hyperfine './my-program', extract mean time. Lower is better."

Read the full file on GitHub · 276 lines

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. 11d ago First seen · 276 lines · 116 tokens per session scan A 754179e98751

Subscribe to this mod's changes

autoresearch is a skill published in the GitHub repository a-tokyo/agent-skills (15 stars, last pushed 1mo ago), licensed MIT. It adds 116 tokens to every session and 2,571 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to autoresearch, differing in 0 lines, and is treated as a copy.

Related

Other skills, from other repositories

writing-plans

Use when you have a spec or requirements for a multi-step task, before touching code.

Bilal140202/the-lord-of-the-skills · 21 tokens

task-reviewer

A structured review process for checking task and verification-plan documents against project requirements. It models requirements, user stories, tasks, contracts, and the evidence needed to show that work is complete.

Haaaiawd/ANWS · 83 tokens

e2e-testing-guide

A guide for writing human-led end-to-end testing documents. End-to-end testing checks a product from the user’s point of view across complete tasks, while manual verification means a person performs the checks.

Haaaiawd/ANWS · 106 tokens

task-planner

A planning method that breaks a system design into a work breakdown structure, meaning a hierarchy of concrete tasks and dependencies. It produces an execution checklist and a separate verification plan in Chinese-language project documents.

Haaaiawd/ANWS · 51 tokens

tdd

Strict test-driven development (Beck-style). Use when writing new code, adding features, or fixing bugs with TDD discipline. TRIGGER when: user asks to use TDD, write tests first, or red/green/refactor. DO NOT TRIGGER when: writing tests after the fact, or user explicitly opts out of TDD.

patforna/auto-task · 72 tokens

gcp-qa

GCP QA and review persona. Critiques and evaluates designs, implementations, and release candidates against acceptance criteria. Owns linting, freshness checks, link validation, and smoke tests. Raises blockers before release. Quality gate authority for both the plugin itself and solutions built with it. Use when the…

jpantsjoha/googlecloud-plugin · 92 tokens