[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6JGKAJXZF7JEEBWXWJBHHDP":3,"skill-same:01M6JGKAJXZF7JEEBWXWJBHHDP":78},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":31,"publishedAt":52,"linkOnly":53,"markdownContent":54,"markdownBody":55,"frontMatter":56,"refName":57,"commitSha":58,"skillDirectory":59,"forksCount":60,"repositoryHtmlUrl":61,"repositoryHomepageUrl":6,"ownerHtmlUrl":62,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":63,"install":64,"resources":77},"01M6JGKAJXZF7JEEBWXWJBHHDP","agentic-topology-design",null,"Choose a multi-agent topology and enforce structured contracts between sub-agents. Use when designing a flow with two or more sub-agents (code-review pipeline, deep research, incident triage, batch refactor), replacing a monolithic prompt agent showing attention dilution, or drafting an ADR for an agentic feature.","muratovv\u002Fai-hats","muratovv","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F7502801?v=4","01M6JGCBN4KEJX012GPJVSH04T","Fedor Muratov","01M6JGCBSNMN7CW3RGTPPE0MN2",3,"MIT","https:\u002F\u002Fgithub.com\u002Fmuratovv\u002Fai-hats\u002Fblob\u002Fmaster\u002Fpackages\u002Fai-hats-library\u002Fsrc\u002Fai_hats_library\u002Fusage\u002Fskills\u002Fagentic-topology-design\u002FSKILL.md",[18,23,26],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000015","视觉设计","SCENARIO","使用场景",{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发",{"categoryNo":27,"categoryName":28,"taxonomyType":29,"taxonomyTypeDesc":30},"01JBSEEDCAT000000000000032","设计师","PROFESSION","适用职业",[32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51],"agentic","agentic topology design","topology","design","use","for","more","feature","attention","deep","incident","enforce","code-review","prompt","choose","adr","pipeline","deep research","code","batch","2026-11-10 10:12:30",false,"---\nname: agentic-topology-design\ndescription: >-\n  Choose a multi-agent topology and enforce structured contracts between\n  sub-agents. Use when designing a flow with two or more sub-agents\n  (code-review pipeline, deep research, incident triage, batch refactor),\n  replacing a monolithic prompt agent showing attention dilution, or\n  drafting an ADR for an agentic feature.\nlicense: MIT\n---\n# Agentic Topology Design\n\nPick the right multi-agent shape for a given task and enforce machine-parseable contracts between sub-agents. Treat sub-agents as microservices with tightly scoped prompts; coordination is the architect's responsibility.\n\n## When to Use\nOwns the *shape* of a multi-agent flow — how many agents, how they hand off,\nwhat contract passes between them. The orthogonal concern is the *trust surface*\nof those agents (where agent output crosses into privileged action) → that's\n**trust-boundary-mapping**. Design the topology here, then map its guardrails\nthere; an agentic feature that touches production needs both.\n\n## Topology Decision Matrix\n\n| Topology | Coordination | Use when | Failure mode it prevents |\n|----------|-------------|---------|---------------------------|\n| **Sequential (Pipeline)** | Output of step N → input of step N+1 | Deterministic order, each step narrows focus (requirements → schema → code → tests) | Hallucinations from over-broad single prompt |\n| **Parallel (Coordinator\u002FWorkers)** | Coordinator fans out independent tasks, aggregates JSON | Latency-critical or independent perspectives (multi-critic review, parallel log scan) | Single-agent code review with hundreds of nitpicks and missed real issues |\n| **Hierarchical Decomposition** | Top agent plans abstract DAG; leaf agents call concrete tools (MCP) | Open-ended research, multi-stage incidents, dynamic scope | Premature commitment to a fixed plan |\n\n## Conventions\n\n- **Structured Outputs Mandate.** Every sub-agent emits JSON Schema-validated output. No free-form natural-language hand-offs between agents — that path leads to brittle parsing and silent corruption.\n- **Tightly-Scoped Sub-Agent Prompts.** Each sub-agent owns one responsibility (search bugs \u002F audit error handling \u002F check security \u002F lint corporate rules). Resist bundling.\n- **Supervisor Owns the DAG.** One coordinator agent forms and mutates the execution graph; workers do not call other workers directly.\n- **Bounded Concurrency.** Cap parallel sub-agents (Brooks\u002FRingelmann projection — see `context-handoff`); coordination overhead grows superlinearly past ~4–6 workers on the same artifact.\n- **Idempotent Re-runs.** Sub-agent contracts must support replay with the same input → same output for cache-friendly recovery.\n\n## Anti-Patterns\n- One mega-prompt that \"routes, formats, queries DB, and styles output\" — instructions interfere; safety bullets get crowded out\n- Free-text contracts (\"write a paragraph the next agent will understand\") — debug it once and you'll never repeat\n- Workers calling workers ad-hoc — the DAG becomes implicit and unmonitorable\n- Picking Hierarchical for a deterministic 4-step flow — over-engineering; Sequential wins\n","# Agentic Topology Design\n\nPick the right multi-agent shape for a given task and enforce machine-parseable contracts between sub-agents. Treat sub-agents as microservices with tightly scoped prompts; coordination is the architect's responsibility.\n\n## When to Use\nOwns the *shape* of a multi-agent flow — how many agents, how they hand off,\nwhat contract passes between them. The orthogonal concern is the *trust surface*\nof those agents (where agent output crosses into privileged action) → that's\n**trust-boundary-mapping**. Design the topology here, then map its guardrails\nthere; an agentic feature that touches production needs both.\n\n## Topology Decision Matrix\n\n| Topology | Coordination | Use when | Failure mode it prevents |\n|----------|-------------|---------|---------------------------|\n| **Sequential (Pipeline)** | Output of step N → input of step N+1 | Deterministic order, each step narrows focus (requirements → schema → code → tests) | Hallucinations from over-broad single prompt |\n| **Parallel (Coordinator\u002FWorkers)** | Coordinator fans out independent tasks, aggregates JSON | Latency-critical or independent perspectives (multi-critic review, parallel log scan) | Single-agent code review with hundreds of nitpicks and missed real issues |\n| **Hierarchical Decomposition** | Top agent plans abstract DAG; leaf agents call concrete tools (MCP) | Open-ended research, multi-stage incidents, dynamic scope | Premature commitment to a fixed plan |\n\n## Conventions\n\n- **Structured Outputs Mandate.** Every sub-agent emits JSON Schema-validated output. No free-form natural-language hand-offs between agents — that path leads to brittle parsing and silent corruption.\n- **Tightly-Scoped Sub-Agent Prompts.** Each sub-agent owns one responsibility (search bugs \u002F audit error handling \u002F check security \u002F lint corporate rules). Resist bundling.\n- **Supervisor Owns the DAG.** One coordinator agent forms and mutates the execution graph; workers do not call other workers directly.\n- **Bounded Concurrency.** Cap parallel sub-agents (Brooks\u002FRingelmann projection — see `context-handoff`); coordination overhead grows superlinearly past ~4–6 workers on the same artifact.\n- **Idempotent Re-runs.** Sub-agent contracts must support replay with the same input → same output for cache-friendly recovery.\n\n## Anti-Patterns\n- One mega-prompt that \"routes, formats, queries DB, and styles output\" — instructions interfere; safety bullets get crowded out\n- Free-text contracts (\"write a paragraph the next agent will understand\") — debug it once and you'll never repeat\n- Workers calling workers ad-hoc — the DAG becomes implicit and unmonitorable\n- Picking Hierarchical for a deterministic 4-step flow — over-engineering; Sequential wins\n",{"name":5,"license":15,"description":7},"master","b4819add93efbaf4d80627118ec34680406b309b","packages\u002Fai-hats-library\u002Fsrc\u002Fai_hats_library\u002Fusage\u002Fskills\u002Fagentic-topology-design",1,"https:\u002F\u002Fgithub.com\u002Fmuratovv\u002Fai-hats","https:\u002F\u002Fgithub.com\u002Fmuratovv","2026-11-05 20:59:10",{"source":8,"skillName":5,"commands":65,"prompt":75,"zipAvailable":76},[66,69,72],{"tool":67,"command":68},"npx","npx skills add muratovv\u002Fai-hats --skill agentic-topology-design",{"tool":70,"command":71},"bunx","bunx skills add muratovv\u002Fai-hats --skill agentic-topology-design",{"tool":73,"command":74},"pnpm","pnpm dlx skills add muratovv\u002Fai-hats --skill agentic-topology-design","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmuratovv\u002Fai-hats\u002Fblob\u002Fmaster\u002Fpackages\u002Fai-hats-library\u002Fsrc\u002Fai_hats_library\u002Fusage\u002Fskills\u002Fagentic-topology-design\u002FSKILL.md",true,[],[79,104,129,148,172,186],{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":52},"01M6PVFB1SM738R2SRSRSF7CY6","tool-evaluation-protocol","Time-bounded protocol for evaluating new tools, libraries, or methodologies before adoption. Use when evaluating a candidate dependency for the project, researching a methodology (e.g. ML\u002FLLM frameworks, testing approaches), or following up on a hype-driven recommendation.","https:\u002F\u002Fgithub.com\u002Fmuratovv\u002Fai-hats\u002Fblob\u002Fmaster\u002Fpackages\u002Fai-hats-library\u002Fsrc\u002Fai_hats_library\u002Fexperimental\u002Fusage\u002Fskills\u002Ftool-evaluation-protocol\u002FSKILL.md",[],[86,87,88,89,90,91,92,93,94,95,36,96,97,98,99,100,101,37,102,103],"protocol","Evaluation","tool","tool evaluation protocol","testing","ml","libraries","adoption","methodology","Tools","dependency","up","frameworks","time","project","recommendation","bounded","candidate",{"skillNo":105,"name":106,"nameZh":6,"description":107,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":108,"categories":109,"keywords":110,"publishedAt":52},"01M6PVF7HATBEX1G7N50E2QT57","systematic-debugging","4-phase bug-fix protocol (evidence, pattern, hypothesis, verify). Use when investigating any error, failure, or unexpected behavior, handling bug reports from users or monitoring, or diagnosing test failures with non-obvious causes.","https:\u002F\u002Fgithub.com\u002Fmuratovv\u002Fai-hats\u002Fblob\u002Fmaster\u002Fpackages\u002Fai-hats-library\u002Fsrc\u002Fai_hats_library\u002Fexperimental\u002Fusage\u002Fskills\u002Fsystematic-debugging\u002FSKILL.md",[],[111,112,113,114,86,115,116,117,118,119,36,120,121,122,123,124,125,126,127,128],"systematic debugging","debugging","systematic","test","monitoring","hypothesis","verify","behavior","fix","evidence","bug","reports","pattern","users","error","with","bug-fix","phase",{"skillNo":130,"name":131,"nameZh":6,"description":132,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":133,"categories":134,"keywords":136,"publishedAt":52},"01M6PVF3MTQN3RPF73AN34856J","skill-template","Canonical template and validation guide for ai-hats skills. Use when creating a new skill, reviewing an existing one for structural compliance, or deciding which pattern fits a new behavior.","https:\u002F\u002Fgithub.com\u002Fmuratovv\u002Fai-hats\u002Fblob\u002Fmaster\u002Fpackages\u002Fai-hats-library\u002Fsrc\u002Fai_hats_library\u002Fexperimental\u002Fusage\u002Fskills\u002Fskill-template\u002FSKILL.md",[135],{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},[137,138,139,140,141,118,142,36,143,123,144,145,146,37,147],"template","skill template","validation","compliance","hats","guide","new skill","structural","canonical","which","Existing",{"skillNo":149,"name":150,"nameZh":6,"description":151,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":152,"categories":153,"keywords":155,"publishedAt":52},"01M6PVF06SCV28QARM5MQPANC9","skill-optimization","Audit and refactor library components to eliminate redundancy and staleness. Use during periodic library maintenance, after adding multiple new skills or rules, when a role's token budget is too high, or after retrospectives identify component issues.","https:\u002F\u002Fgithub.com\u002Fmuratovv\u002Fai-hats\u002Fblob\u002Fmaster\u002Fpackages\u002Fai-hats-library\u002Fsrc\u002Fai_hats_library\u002Fexperimental\u002Fusage\u002Fskills\u002Fskill-optimization\u002FSKILL.md",[154],{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},[156,157,158,159,160,161,162,163,164,165,36,166,167,168,169,170,171],"Optimization","skill optimization","components","component","issues","library","audit","refactor","Budget","Rules","periodic","maintenance","token","token budget","role","high",{"skillNo":173,"name":174,"nameZh":6,"description":175,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":176,"categories":177,"keywords":178,"publishedAt":52},"01M6PVEXM9717016T9MW82HDCT","silent-fallback-gate","PreToolUse hook — refuses an Edit or Write that adds an inert broad except. Use when a Python edit was blocked for a silent fallback.","https:\u002F\u002Fgithub.com\u002Fmuratovv\u002Fai-hats\u002Fblob\u002Fmaster\u002Fpackages\u002Fai-hats-library\u002Fsrc\u002Fai_hats_library\u002Fexperimental\u002Fusage\u002Fskills\u002Fsilent-fallback-gate\u002FSKILL.md",[],[179,180,181,182,183,184,36,185,37],"gate","fallback","silent fallback gate","python","write","edit","hook",{"skillNo":187,"name":188,"nameZh":6,"description":189,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":190,"categories":191,"keywords":193,"publishedAt":52},"01M6PVESM49C5NT7SB7N3F76VS","retro-to-framework","Convert an observation about agent behavior — from a retrospective or from a single incident — into a framework-level change (hook, rule, skill, skill update). Use after a retrospective identifies problems that are not project-specific, after an incident where an agent misapplied a component it had read, when CLAUDE.md band-aids accumulate (more than 3 per-project rules that could be generic), or when the same problem recurs across multiple projects.","https:\u002F\u002Fgithub.com\u002Fmuratovv\u002Fai-hats\u002Fblob\u002Fmaster\u002Fpackages\u002Fai-hats-library\u002Fsrc\u002Fai_hats_library\u002Fexperimental\u002Fusage\u002Fskills\u002Fretro-to-framework\u002FSKILL.md",[192],{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},[194,195,196,197,159,198,199,200,165,118,36,42,201,100,202,203,38,204,205,206],"framework","retro","retro to framework","change","retrospective","projects","convert","across","problem","about","skill update","single","generic"]