sdlc-skills
sourcechecks passing
Agent skill packs for lifecycle work that is easy to fake and hard to verify: false-green tests, mutation gates, incidents, design decisions, release judgment.
markdown
Public Projects
Open-source repositories I maintain. Each badge links to the latest CI run on the default branch; "not recently tested" means the last run is more than 30 days old.
agent-lens
livetests passing
Framework-agnostic, local-first debugger for LLM agents: pause, fork, and compare runs with a numeric verdict.
pythonpypi
falsify
livetests passing
Gate a backtested trading edge before trusting it: multiple-testing, cost-floor and autocorrelation checks in pure-stdlib Python.
pythonpypi
autonomous-investor-writeup
livebuild passing
Case study of the autonomous-investor research platform, published as a static site.
htmlgithub-pages
self-healing-browser-mcp
livetests passing
MCP server that gives agents a real Playwright browser with self-healing locators that recover when selectors drift.
pythonplaywrightmcppypi
playwright-e2e-demo
sourcetests passing
Playwright and TypeScript end-to-end suite: Page Object Model, fixtures, data-driven cases, cross-browser runs in CI.
typescriptplaywright
Questions or feedback?
Found a bug, want to use one of these, or have a question about how it works? Send me a message.