多代理辯證系統的 Critic Agent,負責審查方案並提出挑戰
多代理辯證系統的協調者,負責分析需求並決定三個 Agent 的思考角度
多代理辯證系統的 Perspective Agent A,從指定角度提出解決方案
多代理辯證系統的 Perspective Agent B,從指定角度提出解決方案
多代理辯證系統的 Perspective Agent C,從指定角度提出解決方案
Uses power tools
Uses Bash, Write, or Edit tools
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub chinlung/claude-dev-workflow --plugin multi-agent-debateMulti-agent council orchestration via Claude Code agent teams.
Multi-agent deliberation for AI coding assistants
Convene parallel role-specialized peer agents (dynamic roster, plan card first) to debate a cross-domain decision or audit a codebase in real time. Invoking Claude acts as CEO: convenes, routes peer-DMs, arbitrates deadlocks, writes a one-page decision log.
Multi-agent adversarial review panel — 4-6 AI reviewers debate your code/plans, then a judge delivers a structured verdict with epistemic labels. Bundles plan-review-integrator for applying review findings back into implementation plans.
Complete development workflow system with 6 specialized agents: requirements analysis, code exploration, architecture design, implementation, testing, and quality assurance.
Session 經驗學習系統 - 漸進式保存對話中的有價值模式為 memory 或 skill
High-Precision Multi-Agent development mode - compress single-agent error rate from p to p^4 through epistemic division of labor
Six-phase feature development workflow that enforces strict role separation between OpenSpec (WHAT to build) and Superpowers (HOW to build). Prevents common anti-patterns like modifying specs during code review, patching specs incrementally instead of clean rewriting, and mixing cross-cutting constitution rules into feature specs.
Code review & audit toolkit — routine two-round review commands (/review-branch, /review-pr) sharing path-matched language rule packs, plus quantitative decision frameworks (EV calculation, score-based calibration, STRIDE+CWE classification) and deterministic engineering guarantees (mechanical scope/coverage reconciliation, quoted-code reference anchoring) for high-stakes audits where intuition is insufficient. The toolkit is auto-fix-free and injection-safe by design: it reports and verifies, never rewrites your code — for finding-to-patch work use the claude-security plugin's suggest-patches job.
Collect and synthesize opinions from multiple AI Agents for Claude Code
Consult multiple AI coding agents (Gemini, OpenAI, Grok, Perplexity, plus codex, antigravity, and grok CLIs when installed) to get diverse perspectives on coding problems