[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-ccec14e3-cfd6-416c-bbb3-5564dac24cb0":3,"$fZIu3hxoVbnFP94lWruF-Xrah0bIkg683tYiSrLdc4Vc":43},{"id":4,"title":5,"description":6,"categoryId":7,"moduleId":8,"tags":9,"prompt":10,"icon":11,"source":12,"sourceUrl":13,"authorId":14,"authorName":15,"isPublic":16,"stars":17,"runs":18,"createdAt":19,"updatedAt":19,"module":20,"category":27,"packages":34},"ccec14e3-cfd6-416c-bbb3-5564dac24cb0","multi-agent-brainstorming","使用多个专业代理模拟结构化同行评审流程，以验证设计、揭示隐藏的假设并在实施前识别故障模式。","cat_life_career","mod_other","sickn33,other","---\nname: multi-agent-brainstorming\ndescription: \"Simulate a structured peer-review process using multiple specialized agents to validate designs, surface hidden assumptions, and identify failure modes before implementation.\"\nrisk: unknown\nsource: community\ndate_added: \"2026-02-27\"\n---\n\n# Multi-Agent Brainstorming (Structured Design Review)\n\n## Purpose\n\nTransform a single-agent design into a **robust, review-validated design**\nby simulating a formal peer-review process using multiple constrained agents.\n\nThis skill exists to:\n- surface hidden assumptions\n- identify failure modes early\n- validate non-functional constraints\n- stress-test designs before implementation\n- prevent idea swarm chaos\n\nThis is **not parallel brainstorming**.\nIt is **sequential design review with enforced roles**.\n\n---\n\n## Operating Model\n\n- One agent designs.\n- Other agents review.\n- No agent may exceed its mandate.\n- Creativity is centralized; critique is distributed.\n- Decisions are explicit and logged.\n\nThe process is **gated** and **terminates by design**.\n\n---\n\n## Agent Roles (Non-Negotiable)\n\nEach agent operates under a **hard scope limit**.\n\n### 1️⃣ Primary Designer (Lead Agent)\n\n**Role:**\n- Owns the design\n- Runs the standard `brainstorming` skill\n- Maintains the Decision Log\n\n**May:**\n- Ask clarification questions\n- Propose designs and alternatives\n- Revise designs based on feedback\n\n**May NOT:**\n- Self-approve the final design\n- Ignore reviewer objections\n- Invent requirements post-lock\n\n---\n\n### 2️⃣ Skeptic \u002F Challenger Agent\n\n**Role:**\n- Assume the design will fail\n- Identify weaknesses and risks\n\n**May:**\n- Question assumptions\n- Identify edge cases\n- Highlight ambiguity or overconfidence\n- Flag YAGNI violations\n\n**May NOT:**\n- Propose new features\n- Redesign the system\n- Offer alternative architectures\n\nPrompting guidance:\n> “Assume this design fails in production. Why?”\n\n---\n\n### 3️⃣ Constraint Guardian Agent\n\n**Role:**\n- Enforce non-functional and real-world constraints\n\nFocus areas:\n- performance\n- scalability\n- reliability\n- security & privacy\n- maintainability\n- operational cost\n\n**May:**\n- Reject designs that violate constraints\n- Request clarification of limits\n\n**May NOT:**\n- Debate product goals\n- Suggest feature changes\n- Optimize beyond stated requirements\n\n---\n\n### 4️⃣ User Advocate Agent\n\n**Role:**\n- Represent the end user\n\nFocus areas:\n- cognitive load\n- usability\n- clarity of flows\n- error handling from user perspective\n- mismatch between intent and experience\n\n**May:**\n- Identify confusing or misleading aspects\n- Flag poor defaults or unclear behavior\n\n**May NOT:**\n- Redesign architecture\n- Add features\n- Override stated user goals\n\n---\n\n### 5️⃣ Integrator \u002F Arbiter Agent\n\n**Role:**\n- Resolve conflicts\n- Finalize decisions\n- Enforce exit criteria\n\n**May:**\n- Accept or reject objections\n- Require design revisions\n- Declare the design complete\n\n**May NOT:**\n- Invent new ideas\n- Add requirements\n- Reopen locked decisions without cause\n\n---\n\n## The Process\n\n### Phase 1 — Single-Agent Design\n\n1. Primary Designer runs the **standard `brainstorming` skill**\n2. Understanding Lock is completed and confirmed\n3. Initial design is produced\n4. Decision Log is started\n\nNo other agents participate yet.\n\n---\n\n### Phase 2 — Structured Review Loop\n\nAgents are invoked **one at a time**, in the following order:\n\n1. Skeptic \u002F Challenger\n2. Constraint Guardian\n3. User Advocate\n\nFor each reviewer:\n- Feedback must be explicit and scoped\n- Objections must reference assumptions or decisions\n- No new features may be introduced\n\nPrimary Designer must:\n- Respond to each objection\n- Revise the design if required\n- Update the Decision Log\n\n---\n\n### Phase 3 — Integration & Arbitration\n\nThe Integrator \u002F Arbiter reviews:\n- the final design\n- the Decision Log\n- unresolved objections\n\nThe Arbiter must explicitly decide:\n- which objections are accepted\n- which are rejected (with rationale)\n\n---\n\n## Decision Log (Mandatory Artifact)\n\nThe Decision Log must record:\n\n- Decision made\n- Alternatives considered\n- Objections raised\n- Resolution and rationale\n\nNo design is considered valid without a completed log.\n\n---\n\n## Exit Criteria (Hard Stop)\n\nYou may exit multi-agent brainstorming **only when all are true**:\n\n- Understanding Lock was completed\n- All reviewer agents have been invoked\n- All objections are resolved or explicitly rejected\n- Decision Log is complete\n- Arbiter has declared the design acceptable\n- \nIf any criterion is unmet:\n- Continue review\n- Do NOT proceed to implementation\nIf this skill was invoked by a routing or orchestration layer, you MUST report the final disposition explicitly as one of: APPROVED, REVISE, or REJECT, with a brief rationale.\n---\n\n## Failure Modes This Skill Prevents\n\n- Idea swarm chaos\n- Hallucinated consensus\n- Overconfident single-agent designs\n- Hidden assumptions\n- Premature implementation\n- Endless debate\n\n---\n\n## Key Principles\n\n- One designer, many reviewers\n- Creativity is centralized\n- Critique is constrained\n- Decisions are explicit\n- Process must terminate\n\n---\n\n## Final Reminder\n\nThis skill exists to answer one question with confidence:\n\n> “If this design fails, did we do everything reasonable to catch it early?”\n\nIf the answer is unclear, **do not exit this skill**.\n\n## When to Use\nThis skill is applicable to execute the workflow or actions described in the overview.\n\n## Limitations\n- Use this skill only when the task clearly matches the scope described above.\n- Do not treat the output as a substitute for environment-specific validation, testing, or expert review.\n- Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.\n","","imported","https:\u002F\u002Fgithub.com\u002Fsickn33\u002Fantigravity-awesome-skills","user_system_seed","SkillOPIC",true,78,112,"2026-05-16 13:29:51",{"id":8,"name":21,"slug":22,"icon":23,"description":24,"sort":25,"createdAt":26},"其他","other","mdi-page-next-outline","其他类型Skill",5,"2026-05-16 12:53:40",{"id":7,"name":28,"slug":29,"icon":30,"description":31,"moduleId":8,"sort":32,"skillCount":33,"createdAt":26},"职场发展","career","mdi-briefcase-outline","面试准备、简历优化、职业规划",4,575,[35],{"id":36,"skillId":4,"version":37,"fileName":38,"fileSize":39,"filePath":40,"fileHash":41,"manifest":42,"createdAt":19},"abcb2c3d-6b23-4f4b-9c38-8a494367ed47","1.0.0","multi-agent-brainstorming.zip",2581,"uploads\u002Fskills\u002Fccec14e3-cfd6-416c-bbb3-5564dac24cb0\u002Fmulti-agent-brainstorming.zip","bc353c2727f3628b6b5815972e420876758fe7a4cd2b7766334a483b0cf74190","[{\"path\":\"SKILL.md\",\"isDirectory\":false,\"size\":5733}]",{"code":44,"message":45,"data":46},200,"success",{"items":47,"stats":48,"page":51},[],{"averageRating":49,"totalRatings":49,"ratingCounts":50},0,[49,49,49,49,49],{"limit":52,"offset":49,"hasMore":53,"nextOffset":52,"ratedOnly":16},15,false]