autoresearch
Skill17 starsChanged 24 days agoMITView on GitHub
Run Karpathy-style autoresearch optimization on any content. Generates 50+ variants, scores with a 5-expert simulated panel, evolves winners through multiple rounds, outputs optimized version + full experiment log. Use when optimizing landing pages, email sequences, ad copy, headlines, form pages, CTA text, or any conversion-focused content. Triggers on "optimize this page", "run autoresearch", "score these variants", "A/B test this copy".
105 other projects use a version of this, 2 of them edited.
Adopted by
Most recently taken up by
- matt-riley/agent-skillstheir own version, first seen Nov 15, 2026
- tboome33/obsidian-mcp-routertheir own version, first seen Nov 15, 2026
- AndyMelnykov/obsidian-claudetheir own version, first seen Nov 15, 2026
- thisis-romar/opendeck-factorytheir own version, first seen Nov 15, 2026
- arbrain-ai/arbraintheir own version, first seen Nov 15, 2026
- darkroomengineering/cc-settingstheir own version, first seen Nov 12, 2026
Used with
Projects with it also use aeon-update, auto-workflow and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
aaronjmarsSkillaeon-update
Pull framework updates from the upstream Aeon repo into this instance - 3-way merges canon's new commits into a PR, never clobbering operator config.
Also in 4 projects that use this
aaronjmarsSkillauto-workflow
Two-mode aeon.yml workflow builder - analyze inspects URLs and emits a tiered, signal-verified skill-enablement plan plus an aeon.yml diff; enable flips slugs to enabled:true and opens a PR.
Also in 4 projects that use this
- +10 more from aaronjmars
Other versions105 projects use a version of this.
Word for word · 0
None found.
Edited · 2
- ericosiu/ai-marketing-skills★ 3.8k22% same
- admdanielpaulino/claude-codes★ 122% same
Wrote their own version · 103
- github/awesome-copilot★ 41ksame name
- Yeachan-Heo/oh-my-claudecode★ 41ksame name
- Yeachan-Heo/oh-my-claude-sisyphus★ 41ksame name
- Yeachan-Heo/oh-my-codex★ 34ksame name
- AgriciDaniel/claude-obsidian★ 16ksame name
- Orchestra-Research/AI-research-SKILLs★ 13ksame name
- zechenzhangAGI/AI-research-SKILLs★ 13ksame name
- Orchestra-Research/AI-Research-SKILLs★ 12ksame name
- getcompanion-ai/feynman★ 10ksame name
- advaitpaliwal/feynman★ 10ksame name
93 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository38 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- adversarial-claims-reviewerSkill
- browser-testing-with-devtoolsSkill
- codebase-cost-estimatorSkill
- code-review-and-qualitySkill
- content-opsSkill
- content-pipelineSkill
- conversion-opsSkill
- data-model-documentationSkill
- data-model-verificationSkill
- deployment-pipelinesSkill
- findings-ledgerSkill
- game-balancerSkill
- game-systems-designerSkill
- godot-engineerSkill
- growth-engineSkill
- iap-managerSkill
- library-investigatorSkill
- marketing-shaperSkill
- memory-extractionSkill
- outbound-engineSkill
- phaser-engineerSkill
- planning-and-task-breakdownSkill
- prompt-shaperSkill
- release-managerSkill
- revenue-intelligenceSkill
- rust-engineerSkill
- security-engineeringSkill
- securitySkill
- seo-opsSkill
- session-stateSkill
- skill-library-reviewSkill
- telemetrySkill
- typescript-analyticsSkill
- typescript-data-engineeringSkill
- typescript-testing-backendSkill
- typescript-testing-frontendSkill
- web3-smart-contract-engineeringSkill
The file
We split this skill into the rules it gives an agent
7 rules. 0 are common, 7 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 101 edited versions we read keep these.
7
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 19
Every run outputs 3 files:
- Distinctive: line 31
Run up to 3 rounds per element.
- Distinctive: line 33
Write 3 output files: optimized content, experiments JSON, optimization report.
- Distinctive: line 49
< 70: Don't ship. Something fundamental is broken.
- Distinctive: line 57
Never call the API once per variant.
- Distinctive: line 58
Don't over-optimize for one dimension — a 95 clarity + 45 trust is a misleading overall score.
- Distinctive: line 59
Don't run more than 5 rounds.
“Common” is measured: kept by at least 6 in 10 of the 101 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Autoresearch Skill
Karpathy-style optimization loops for any conversion-focused content. No traffic needed. Simulated expert panel. Minutes, not weeks.
The sequence: Run autoresearch FIRST to hit 85+ simulated score. Then deploy. Then validate with real traffic.
What You'll Produce
Every run outputs 3 files:
| File | Purpose |
{name}-optimized.{ext} | The winning optimized content |
data/{name}-experiments.json | Full experiment log — all variants + all scores |
data/{name}-optimization-report.md | Human-readable summary with winner rationale |
Core rules
- Intake: parse content type from file extension (HTML → landing/form page; markdown → email/ad copy). List all optimizable elements and confirm with user before starting.
- Score every variant against the 5-expert panel in a single batched API call per round — never one call per variant.
- Run up to 3 rounds per element; stop early when top variant hits the minimum score threshold (default: 80).
- Cross-breed only after each element has its own winner — premature cross-breeding creates incoherent combinations.
- Write 3 output files: optimized content, experiments JSON, optimization report.
- Report back: winning score, biggest score jump, top 2 runner-ups, file paths, next step.
User Options
| Option | Default | Description |
elements | all | Which elements to optimize |
variants_per_round | 10 | How many variants per round |
min_score | 80 | Stop when this score is hit |
rounds | 3 | Max rounds before stopping |
auto_apply | false | Overwrite source file with winners |
content_type | auto-detect | Force a content type if auto-detect is wrong |
Quality Gates
- < 70: Don't ship. Something fundamental is broken.
- 70-79: Marginal. One more round targeting the lowest-scoring dimension.
- 80-84: Good. Shippable. Validate with real traffic.
- 85-89: Strong. Ship with confidence.
- 90+: Rare. Ship immediately.
Anti-Patterns
- Never call the API once per variant. Always batch. A 10-variant round = 1 call.
- Don't over-optimize for one dimension — a 95 clarity + 45 trust is a misleading overall score.
- Don't run more than 5 rounds. If not hitting 80 after 3, the problem is strategic, not tactical.
References
- references/round-structure.md — round-by-round procedure and 5-expert panel definitions
- references/content-types.md — elements and score dimensions per content type (landing pages, email, ads, forms)
- references/execution-protocol.md — 6-step execution protocol with JSON schema
Related skills
- content-ops — quality scoring gate; run after autoresearch optimization
- conversion-ops — audits a LIVE landing-page URL across CRO dimensions; use it instead when scoring a live page rather than optimizing copy variants pre-launch
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

