
Claude Skills by startmeupai
github.com/startmeupaiAudit authentication, resource scope, IDOR, and apply-time authorization.
Audit AI trust boundaries, wrapper use, strict outputs, and apply-time controls.
Maintain Cloudflare Worker configuration, routes, bindings, builds, and validation.
Audit dependency advisories, lockfile integrity, lifecycle scripts, and parser risk.
Create concise, structured, link-valid, Markdown-compliant engineering documentation.
Sweep feature claims across access profiles, fix defects, and retain durable coverage.
Deliver a scoped feature across each layer the project has, with authorization, tests, and evidence.
Create and diagnose secure, reproducible GitHub Actions workflows.
Close browser-observable plan gates and preserve fresh, profile-correct evidence.
Execute approved plans continuously with truthful stage-end writeback.
Audit and correct implementation plans until blocking findings are resolved.
Research a repository and, only when requested, produce an evidence-backed staged implementation plan.
Create, debug, and maintain deterministic Playwright browser journeys.
Implement resource-scoped roles, membership, guards, and runtime authorization proof.
Modernize code while preserving externally observable behavior.
Audit source, history, logs, and errors for credential or sensitive-data exposure.
Validate persona, skill, rules, stack-pack, catalog, and runtime-surface consistency.
Capture a redacted step-by-step browser evidence bundle and manifest.
Create deterministic unit, component, integration, and contract tests.
Recreate an approved visual reference with accessible, token-driven responsive UI.
Verify authenticated small-screen routes and interactions with browser evidence.
Compare implemented UI with approved references using structural and visual evidence.
Audit uploads, extraction, and server fetches for SSRF and resource-exhaustion risk.
Build, harden, and validate Dockerfiles, ignore files, and local compose configuration.
Deliver a scoped Python feature with typed boundaries, validated input, explicit layers, and pytest proof.
Create deterministic pytest unit, integration, and property tests with fixtures, parametrization, and isolated side effects.
Write Supabase migrations, RLS policies, edge functions, and types with local-first proof.
Implement TypeScript behavior with strict types, schema-validated boundaries, and exhaustive domain models.
Write deterministic, type-checked TypeScript unit, component, and integration tests with typed fixtures and fakes.
Turn local persona and skill improvements in an adopting repository into a sanitized, reviewed pull request to the swe-agents test branch.
Update an adopting repository's installed personas, skills, and rules from its swe-agents source, three-way merging local edits.