webDevYuri/claude-engineering-runtime

Claude Code Multi-Agent Runtime: a drop-in, Claude Code–native harness that gives a project a multi-agent workflow (one Orchestrator + worker subagents) to plan, implement, design, QA, and document code with AI-maintained memory and living contracts. Target users: developer teams using Claude Code who want a governed automation pipeline. Key features: single Orchestrator that controls task state and memory; strict lane boundaries for planner/backend/frontend/ui-ux/qa/docs/devops; self-updating memory in .claude/memory/ and living contracts in .claude/contracts/; kanban-style task folders (queued/active/blocked/review/completed); built-in commands (/initialize-project, /plan-feature, /implement-feature, /audit-ui, /run-qa, /update-memory); vendored Impeccable design skill for deterministic UI audits and LLM critique.

このRepoのdescription / GitHub topics / READMEのAI要約に出現したAI関連キーワードをカテゴリ別に表示。各バッジは該当カテゴリの詳細ランキングへリンクします。