A loop-engineering system for AI coding agents, bundled as one Claude Code plugin. Converge → specify → implement: parallel TDD development with dual-ring convergence gates, adversarial cross-source review, source-grounded citation verification, and prior-art collision search — honest coverage at every gate, never a silent green.
tdd code-quality ai-agents prior-art-search agentic-ai agent-skills claude-code ai-coding-agent parallel-development spec-driven-development adversarial-review loop-engineering primary-source-verification convergent-loop architecture-contract cross-source-review
-
Updated
Aug 12, 2026 - Python