[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6NV6GDY4KCDKVZMEEQ1E6RH":3,"skill-same:01M6NV6GDY4KCDKVZMEEQ1E6RH":66},{"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":18,"publishedAt":39,"linkOnly":40,"markdownContent":41,"markdownBody":42,"frontMatter":43,"refName":44,"commitSha":45,"skillDirectory":46,"forksCount":14,"repositoryHtmlUrl":47,"repositoryHomepageUrl":6,"ownerHtmlUrl":48,"ownerBlogUrl":49,"ownerDescription":50,"updatedAt":51,"install":52,"resources":65},"01M6NV6GDY4KCDKVZMEEQ1E6RH","layered-recall",null,"4-layer memory recall system. Layer 1 (identity) always loaded, Layer 2 (critical facts) per-project, Layer 3 (room recall) on-demand, Layer 4 (deep search) when needed. Progressive context loading for token efficiency.","rubicanjr\u002FFinCognis","rubicanjr","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F234912519?v=4","01M6NT156NND8J474G34SB3923","Rubi","01M6NT15B70RWXE1N3YV0BZY8A",0,"MIT","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Flayered-recall\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38],"layered recall","recall","layered","critical","token efficiency","for","facts","layer","deep search","loading","memory recall","memory","token","project","deep","efficiency","context","system","search","identity","2026-11-10 10:12:30",false,"---\nname: layered-recall\ndescription: \"4-layer memory recall system. Layer 1 (identity) always loaded, Layer 2 (critical facts) per-project, Layer 3 (room recall) on-demand, Layer 4 (deep search) when needed. Progressive context loading for token efficiency.\"\n---\n\n# Layered Recall\n\n4-layer progressive memory system. Each layer adds more context only when needed, saving tokens while ensuring nothing important is missed.\n\n## The 4 Layers\n\n```\nLayer 1: Identity (always loaded, ~200 tokens)\n   Who is the user? What are their preferences?\n\nLayer 2: Critical Facts (per-project, ~500 tokens)\n   Hard constraints, active decisions, blockers\n\nLayer 3: Room Recall (on-demand, ~1-2K tokens)\n   Relevant memories for current task domain\n\nLayer 4: Deep Search (when needed, ~2-5K tokens)\n   Full semantic search across all memories\n```\n\n## Layer Details\n\n### Layer 1: Identity (~200 tokens, ALWAYS loaded)\n\nLoaded at every session start. Contains:\n- User preferences (language, style, autonomy level)\n- Global constraints (no emojis, Turkish responses, etc.)\n- Tool preferences (which editors, which terminal)\n\n**Source:** `~\u002F.claude\u002Fprojects\u002F*\u002Fmemory\u002Fuser_*.md`\n\n### Layer 2: Critical Facts (~500 tokens, per-project)\n\nLoaded when entering a project directory. Contains:\n- Active architectural decisions\n- Known blockers and constraints\n- Current sprint\u002Fmilestone goals\n- Tech stack and versions\n\n**Source:** `~\u002F.claude\u002Fprojects\u002F*\u002Fmemory\u002Fproject_*.md` + `thoughts\u002FCONTEXT.md`\n\n### Layer 3: Room Recall (~1-2K tokens, on-demand)\n\nLoaded when task domain is detected (auth, database, deploy, etc.). Contains:\n- Previous decisions in this domain\n- Past errors and fixes\n- Patterns that worked\n- Patterns that failed\n\n**Source:** Memory palace rooms + `mature-instincts.json` filtered by domain\n\n**Trigger:** Intent classifier detects domain (e.g., \"fix the login bug\" -> room: authentication)\n\n### Layer 4: Deep Search (~2-5K tokens, explicit)\n\nOnly loaded when explicitly needed or when Layers 1-3 don't have enough context. Contains:\n- Full semantic search results\n- Cross-project pattern matches\n- Historical error resolutions\n- Archived decisions\n\n**Source:** PostgreSQL vector search + palace cross-wing search\n\n**Trigger:** Agent explicitly queries, or user asks \"have we done this before?\"\n\n## Recall Flow\n\n```\nSession Start\n  -> Load Layer 1 (identity)\n  -> Detect project -> Load Layer 2 (facts)\n  -> User sends prompt\n  -> Classify intent\u002Fdomain -> Load Layer 3 (room)\n  -> If insufficient context -> Load Layer 4 (deep)\n```\n\n## Token Budget\n\n| Layer | Tokens | When |\n|-------|--------|------|\n| L1 | ~200 | Always |\n| L2 | ~500 | Per project |\n| L3 | ~1-2K | Per task domain |\n| L4 | ~2-5K | On demand |\n| **Total max** | **~8K** | Worst case |\n\nvs. loading everything: ~30-50K tokens\n\n**Savings: 4-6x token reduction**\n\n## Integration\n\n### With Existing Hooks\n- `instinct-loader` -> feeds Layer 2 and Layer 3\n- `smart-memory-recall` -> implements Layer 3 scoring\n- `intent-classifier` -> triggers Layer 3 room selection\n- `graph-indexer` -> powers Layer 4 deep search\n\n### With Memory Palace\n- Layer 2 pulls from palace wing index\n- Layer 3 pulls from palace room drawers\n- Layer 4 searches across all wings\n\n### With Agents\n- Agents inherit parent's Layer 1-2 context\n- Each agent can request Layer 3-4 for their domain\n- Agent memories feed back into palace for future recall\n","# Layered Recall\n\n4-layer progressive memory system. Each layer adds more context only when needed, saving tokens while ensuring nothing important is missed.\n\n## The 4 Layers\n\n```\nLayer 1: Identity (always loaded, ~200 tokens)\n   Who is the user? What are their preferences?\n\nLayer 2: Critical Facts (per-project, ~500 tokens)\n   Hard constraints, active decisions, blockers\n\nLayer 3: Room Recall (on-demand, ~1-2K tokens)\n   Relevant memories for current task domain\n\nLayer 4: Deep Search (when needed, ~2-5K tokens)\n   Full semantic search across all memories\n```\n\n## Layer Details\n\n### Layer 1: Identity (~200 tokens, ALWAYS loaded)\n\nLoaded at every session start. Contains:\n- User preferences (language, style, autonomy level)\n- Global constraints (no emojis, Turkish responses, etc.)\n- Tool preferences (which editors, which terminal)\n\n**Source:** `~\u002F.claude\u002Fprojects\u002F*\u002Fmemory\u002Fuser_*.md`\n\n### Layer 2: Critical Facts (~500 tokens, per-project)\n\nLoaded when entering a project directory. Contains:\n- Active architectural decisions\n- Known blockers and constraints\n- Current sprint\u002Fmilestone goals\n- Tech stack and versions\n\n**Source:** `~\u002F.claude\u002Fprojects\u002F*\u002Fmemory\u002Fproject_*.md` + `thoughts\u002FCONTEXT.md`\n\n### Layer 3: Room Recall (~1-2K tokens, on-demand)\n\nLoaded when task domain is detected (auth, database, deploy, etc.). Contains:\n- Previous decisions in this domain\n- Past errors and fixes\n- Patterns that worked\n- Patterns that failed\n\n**Source:** Memory palace rooms + `mature-instincts.json` filtered by domain\n\n**Trigger:** Intent classifier detects domain (e.g., \"fix the login bug\" -> room: authentication)\n\n### Layer 4: Deep Search (~2-5K tokens, explicit)\n\nOnly loaded when explicitly needed or when Layers 1-3 don't have enough context. Contains:\n- Full semantic search results\n- Cross-project pattern matches\n- Historical error resolutions\n- Archived decisions\n\n**Source:** PostgreSQL vector search + palace cross-wing search\n\n**Trigger:** Agent explicitly queries, or user asks \"have we done this before?\"\n\n## Recall Flow\n\n```\nSession Start\n  -> Load Layer 1 (identity)\n  -> Detect project -> Load Layer 2 (facts)\n  -> User sends prompt\n  -> Classify intent\u002Fdomain -> Load Layer 3 (room)\n  -> If insufficient context -> Load Layer 4 (deep)\n```\n\n## Token Budget\n\n| Layer | Tokens | When |\n|-------|--------|------|\n| L1 | ~200 | Always |\n| L2 | ~500 | Per project |\n| L3 | ~1-2K | Per task domain |\n| L4 | ~2-5K | On demand |\n| **Total max** | **~8K** | Worst case |\n\nvs. loading everything: ~30-50K tokens\n\n**Savings: 4-6x token reduction**\n\n## Integration\n\n### With Existing Hooks\n- `instinct-loader` -> feeds Layer 2 and Layer 3\n- `smart-memory-recall` -> implements Layer 3 scoring\n- `intent-classifier` -> triggers Layer 3 room selection\n- `graph-indexer` -> powers Layer 4 deep search\n\n### With Memory Palace\n- Layer 2 pulls from palace wing index\n- Layer 3 pulls from palace room drawers\n- Layer 4 searches across all wings\n\n### With Agents\n- Agents inherit parent's Layer 1-2 context\n- Each agent can request Layer 3-4 for their domain\n- Agent memories feed back into palace for future recall\n",{"name":5,"description":7},"main","70e1c350500ee750987420bbae48837afa906b7e","skills\u002Flayered-recall","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis","https:\u002F\u002Fgithub.com\u002Frubicanjr","linktr.ee\u002Ffinanskodu","Industrial Engineer & CIO @FinansKodu. Architecting AI-driven financial agents and autonomous workflows. 🏗️ Building systems, not just prompts.","2026-06-09 13:50:15",{"source":8,"skillName":5,"commands":53,"prompt":63,"zipAvailable":64},[54,57,60],{"tool":55,"command":56},"npx","npx skills add rubicanjr\u002FFinCognis --skill layered-recall",{"tool":58,"command":59},"bunx","bunx skills add rubicanjr\u002FFinCognis --skill layered-recall",{"tool":61,"command":62},"pnpm","pnpm dlx skills add rubicanjr\u002FFinCognis --skill layered-recall","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Flayered-recall\u002FSKILL.md",true,[],[67,88,97,117,146,164],{"skillNo":68,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":71,"categories":72,"keywords":78,"publishedAt":39},"01M6NWS9DAXJTADT4T5MB49X8W","workflow-router","Goal-based workflow orchestration - routes tasks to specialist agents based on user goals","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fworkflow-router\u002FSKILL.md",[73],{"categoryNo":74,"categoryName":75,"taxonomyType":76,"taxonomyTypeDesc":77},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[79,80,81,82,83,84,85,86,87],"workflow router","workflow","router","orchestration","goals","goal","user","workflow orchestration","tasks",{"skillNo":89,"name":90,"nameZh":6,"description":91,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":92,"categories":93,"keywords":94,"publishedAt":39},"01M6NWS3RETSC0SHP16HP66E14","wiring","Wiring Verification","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fwiring\u002FSKILL.md",[],[90,95,96],"verification","false",{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":106,"publishedAt":39},"01M6NWS0XW9CVQZBDA1EGW0TWC","websocket-patterns","Connection management, room patterns, reconnection strategies, message buffering, and binary protocol design.","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fwebsocket-patterns\u002Fskill.md",[103],{"categoryNo":104,"categoryName":105,"taxonomyType":76,"taxonomyTypeDesc":77},"01JBSEEDCAT000000000000015","视觉设计",[107,108,109,110,111,112,113,114,115,116],"websocket","patterns","websocket patterns","design","protocol","management","message","strategies","binary","connection",{"skillNo":118,"name":119,"nameZh":6,"description":120,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":121,"categories":122,"keywords":127,"publishedAt":39},"01M6NWRDQDX9ZNETAWRRMQWV52","visual-verdict","Screenshot comparison QA for frontend development. Takes a screenshot of the current implementation, scores it across multiple visual dimensions, and returns a structured PASS\u002FREVISE\u002FFAIL verdict with concrete fixes. Use when implementing UI from a design reference or verifying visual correctness.","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fvisual-verdict\u002FSKILL.md",[123,126],{"categoryNo":124,"categoryName":125,"taxonomyType":76,"taxonomyTypeDesc":77},"01JBSEEDCAT000000000000009","前端开发",{"categoryNo":74,"categoryName":75,"taxonomyType":76,"taxonomyTypeDesc":77},[128,129,130,131,132,133,134,135,136,24,137,138,139,110,140,141,142,143,144,145],"visual","verdict","visual verdict","ui","across","revise","comparison","correctness","implementing","fixes","dimensions","frontend","development","qa","screenshot","returns","use","frontend development",{"skillNo":147,"name":148,"nameZh":6,"description":149,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":150,"categories":151,"keywords":152,"publishedAt":39},"01M6NWR8HSNRS7MNDK2J02QYMN","verification-loop","Comprehensive verification system covering build, types, lint, tests, security, and diff review before a PR.","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fverification-loop\u002FSKILL.md",[],[153,154,95,155,156,157,158,36,159,160,161,162,163],"verification loop","loop","lint","review","security","pr","diff","build","tests","diff review","types",{"skillNo":165,"name":166,"nameZh":6,"description":167,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":168,"categories":169,"keywords":170,"publishedAt":39},"01M6NWR4HHSFR0VR5EXFRWA747","vector-db-patterns","Embedding strategies, ANN algorithms, hybrid search, RAG chunking strategies, and reranking for semantic search and retrieval.","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fvector-db-patterns\u002Fskill.md",[],[171,172,173,174,108,175,37,176,177,114,178,179,180,181,182,24,183,184],"vector db patterns","vector db","db","vector","semantic search","embedding strategies","algorithms","chunking","semantic","retrieval","embedding","Hybrid Search","hybrid","rag"]