Stop agents from calling buggy code "done": a Claude Code plugin where separate agents test, review and try to break every story, and each bug leaves a check. By Leopoldo Simini.
Claude Agent Skill that audits any SaaS for production readiness (security, scalability, reliability, testing, observability, compliance). Reports first, fixes only after you approve. No scripts, works with any stack.
Don't let your claude session limits break your flow - defer tasks for the next session, CC starts working on them as soon as it can, come back online, read your reports and keep building!
Audit what your coding agents silently ignore: Claude Code's CLAUDE.md, skills, sub-agents, hooks, settings, permissions and MCP, plus Cursor rules, Copilot instructions and AGENTS.md. Python, no dependencies, no network.
Agent harness experiment in Python. Several architectures for turning a markdown spec into a codebase, from a multi-agent pipeline to a spec-compiled task DAG with token-budgeted context and per-stage evals.
Let Claude Code, Codex & Gemini CLI agents talk to each other — a local MCP message bus for multi-agent collaboration, cross-model code review, and fewer hallucinations.