
Autoresearch
SkillRuns an autonomous experiment loop that edits code, measures a metric and keeps wins
- Open source
- 41.3K GitHub stars
- Developer tools

Runs an autonomous experiment loop that edits code, measures a metric and keeps wins

Walks through setting up an AWS CDK Python project from prerequisites to first deploy

Gives CloudWatch query patterns and a decision tree for triaging AWS incidents

Diagnoses one AWS resource's health from CloudWatch data and drafts a remediation plan

Reviews AWS infrastructure code against the six Well-Architected pillars and files issues

Gives the agent az acr command knowledge for Azure Container Registry

Runs Bicep build, what-if and permission checks before an Azure deployment

Guides agents in building, migrating and reviewing azd projects with safety guardrails

Lets the agent run Azure DevOps projects, repos, pipelines, boards and PRs from the CLI

Scaffolds, validates, tests and deploys Azure Logic Apps Standard via az logicapp

Diagnoses an Azure resource from logs and telemetry and drafts a tiered remediation plan

Runs the SWA CLI to set up, test locally and deploy Azure Static Web Apps

Helps the agent write and debug Windows .bat and .cmd scripts with cmd.exe

Adds start, progress, failure and summary output to batch scripts, behaviour unchanged

Guides refactoring React components away from boolean props toward compound components

Rule set for fast React Native and Expo apps covering lists, animation and native UI

Deploys and manages Vercel projects through the CLI using an access token, not login

Audits a deployed Vercel project for cost and speed using metrics before reading any code

Reviews UI code against Vercel's Web Interface Guidelines and lists findings as file:line

Checks docs and prose against Vercel's writing guidelines and reports file:line findings

Deploys a project to Vercel as a preview, choosing git push, CLI or a no-login fallback

70 React and Next.js performance rules ranked by impact, from waterfalls to bundle size

Implements React View Transition animations for navigation, shared elements and lists

Runs Lighthouse across a whole site locally with Unlighthouse and aggregates scores

Audits and fixes how ready a site is for AI browsing agents, from llms.txt to WebMCP

Snapshots on-page SEO elements and diffs them later to catch regressions after a deploy

Guides ASP.NET Core work by picking the right app model and loading only needed references

Implements Figma frames as code through the Figma MCP server in a fixed fetch order

Maps unmapped Figma components to matching code components using Code Connect

Writes project-specific Figma-to-code rules into CLAUDE.md, AGENTS.md or Cursor rules