autoresearch
Autoresearch automates large-scale SKILL.md improvement by spawning multiple variants per generation, evaluating each through a programmatic harness without browser overhead, and scoring with triple-run averaging to identify the strongest performer. The engine iterates autonomously, committing improvements to git and halting when scores plateau or oscillate.
Autoresearch runs parallel SKILL.md optimization, generating and testing 10+ variants per generation with programmatic scoring.
AI-generated summary based on this skill's SKILL.md
Decision gist · record as of 2026-07-27
Autoresearch runs parallel SKILL.md optimization, generating and testing 10+ variants per generation with programmatic scoring. Autoresearch automates large-scale SKILL.md improvement by spawning multiple variants per generation, evaluating each through a programmatic harness without browser overhead, and scoring with triple-run averaging to identify the strongest performer. The engine iterates autonomously, committing improvements to git and halting when scores plateau or oscillate.
Use it when
- Autoresearch generates autonomous variants of your SKILL.md file in parallel.
- Yes.
Verify before relying
Read SKILL.md below before installing (1 file). Open directory: indexed for reading, not audited.
Install
popmechanic/VibesOS/autoresearch · repository language: HTML
Open directory. Skills are indexed for reading, not audited. Review a skill's body before installing it.
Frequently asked questions
AI-generated answers based on this skill's SKILL.md and metadata
What does autoresearch skill do?
Autoresearch automates large-scale SKILL.md improvement by spawning multiple variants per generation, evaluating each through a programmatic harness without browser overhead, and scoring with triple-run averaging to identify the strongest performer. The engine iterates autonomously, committing improvements to git and halting when scores plateau or oscillate.
How does autoresearch run parallel SKILL.md optimization?
Autoresearch generates autonomous variants of your SKILL.md file in parallel, testing each variant through a programmatic evaluation harness that requires no browser dependencies. It scores fitness across all variants using triple-run averaging to smooth results, then selects the strongest performer for the next generation of mutation and testing.
Can autoresearch improve skill quality at scale through multi-generation mutation?
Yes. Autoresearch improves skill quality at scale by running multiple generations of mutation and scoring. Each generation spawns new variants from the best prior performer, evaluates them programmatically, and advances only those that increase fitness. This iterative process continues until scores plateau or oscillate, signaling convergence.
Does autoresearch require a browser for skill evaluation?
No. Autoresearch executes programmatic skill evaluation without browser dependencies, making it lightweight and suitable for CI/CD pipelines and headless environments. The evaluation harness runs directly against your SKILL.md variants without spawning a browser instance.
How does autoresearch analyze skill fitness across variants?
Autoresearch analyzes skill fitness across variants with triple-run averaging, running each variant three times and averaging the scores to reduce noise. This approach identifies genuinely stronger performers rather than rewarding lucky single runs, ensuring stable progression across generations of optimization.
What license does autoresearch use?
Autoresearch is released under the MIT license, permitting free use, modification, and distribution in both open-source and commercial projects with minimal restrictions.
SKILL.md
Rendered from the published skill. Quoted content, verbatim.
Parallel Autoresearch Engine
> Plan mode: This skill is ONE plan step: "Invoke /vibes:autoresearch". Do not decompose.
Prerequisites Check
```bash VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}" echo "Checking autoresearch prerequisites..." test -f "$VIBES_ROOT/scripts/eval-ssr-check.ts" && echo "✓ Tier 1.5 SSR check" || echo "✗ Missing eval-ssr-check.ts" test -f "$VIBES_ROOT/scripts/eval-harness.ts" && echo "✓ Tier 2 harness" || echo "✗ Missing eval-harness.ts" test -f "$VIBES_ROOT/scripts/eval-parallel.ts" && echo "✓ Orchestrator" || echo "✗ Missing eval-parallel.ts" test -f "$VIBES_ROOT/scripts/eval-scoring.ts" && echo "✓
(truncated - see the full file via the links below)
File tree — 1 file
vibes-desktop/build/stable-macos-arm64/VibesOS.app/Contents/Resources/vibes-plugin/skills/autoresearch/SKILL.md
Let your AI agent find skills like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 56,283 agent skills by what they can do, searchable in plain language.
wish › “Run parallel SKILL.md optimization with autonomous variant generation and testing”
Give your agent the search over MCP, or paste the wish link into any chat. No install? Search from any chat →
Related skills
Vibes is a self-contained app generator that creates React web applications powered by TinyBase for reactive, real-time data management. Choose between a browser-based editor with live preview or terminal-driven generation, then deploy your app as a standalone HTML file. Ideal for rapid prototyping, single-page applications, and collaborative tools.
Launch is a self-contained SaaS pipeline that generates a complete app, adds authentication and billing, and deploys it live—all in one invocation. Choose between a visual editor with live preview or terminal-based generation, then let the skill handle setup, configuration, and deployment automatically.
Riff creates multiple app variations simultaneously, each representing a different interpretation rather than just visual tweaks. Gather your requirements once, then generate and rank all variations in parallel to find the strongest concept.
Sell transforms an existing Vibes app into a production-ready SaaS platform with multi-tenant architecture, automatic Pocket ID authentication, and subscription controls. The skill generates a unified HTML file that routes traffic by subdomain—landing page, tenant apps, and admin dashboard all from a single deployment.
This skill automates deployment of Vibes apps to Cloudflare Workers using the Deploy API, eliminating the need for wrangler CLI or manual infrastructure setup. It manages browser-based authentication on first run, caches credentials for future deploys, and bundles static assets automatically. Optionally configure OpenRouter API keys to enable AI features in your deployed app.
Factory transforms a Vibes app into a revenue-generating SaaS by automating subscription setup, AI token billing, and Vibe Token economics for partner rewards. It handles Stripe integration, deploys to Cloudflare Workers, and manages token-backed revenue sharing through a unified configuration pipeline.
More skills design (MIT) · lab:autoresearch (MIT)