Building open-source tools for AI coding agents at App Builders Club @app-builders-club. 8+ years in tech. I build tools where specs and test cases live as plain files in git - and AI agents read, write, and verify them.
QA tooling for the AI-agent era:
- Kensa IDE - the desktop IDE for QA engineers. Test cases as Markdown in your own git repo, local-first, edited by your own terminal AI agent. Ships with a 30-command CLI built for agents, plus Blueprints for node-graph automation. Windows · macOS.
- QA_agents - the Kensa-QA plugin. Spawns a full squad of expert agents (test-lead, ISTQB engineers, plus risk / API / web / mobile / security / performance specialists). Point them at a ticket, get committed test cases back.
Document-Driven Development tools @app-builders-club:
- dev-team-agents - turns one Claude Code session into a coordinated dev team. An Orchestrator dispatches specialist agents (Analyst, Architect, Developer, QA, Reviewer, Debugger…) that hand off through a file bus, with a plan-review stage before any code and an optional Codex second reviewer.
- design-builder - design plugin for Claude Code. Production-grade interfaces on web and iOS.
- designLib-mcp - curated design knowledge catalog served over MCP. Drop-in source of truth for any AI coding agent.
- mvp-builder - Document-Driven Development framework for Claude Code. Structured specs, TDD cycles, self-review, session memory.
→ More tools & the methodology behind them at app-builders.club
- Shipping Kensa IDE + the QA_agents plugin
- Automating a Flutter banking app at scale
- Hundreds of hours deep into Claude Code workflows