[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-82640":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":16,"stars7d":17,"stars30d":18,"stars90d":15,"forks30d":15,"starsTrendScore":17,"compositeScore":19,"rankGlobal":10,"rankLanguage":10,"license":20,"archived":21,"fork":21,"defaultBranch":22,"hasWiki":23,"hasPages":21,"topics":24,"createdAt":10,"pushedAt":10,"updatedAt":33,"readmeContent":34,"aiSummary":35,"trendingCount":15,"starSnapshotCount":15,"syncStatus":16,"lastSyncTime":36,"discoverSource":37},82640,"third-brain-v5-skills","Mark393295827\u002Fthird-brain-v5-skills","Mark393295827","agent skills","https:\u002F\u002Fgithub.com\u002FMark393295827\u002Fthird-brain-v5-skills\u002Ftree\u002Fmaster",null,"HTML",114,17,1,0,2,6,7,51.47,"MIT License",false,"main",true,[25,26,27,28,29,30,31,32,5],"agentic","ai-agents","claude-skills","codex","cursor-rules","knowledge-management","obsidian","second-brain","2026-06-12 04:01:38","# Third Brain V5 Skills\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"assets\u002Fthird-brain-v5-system-architecture.png\" alt=\"Third Brain V5 system architecture: LLM, Skills, Obsidian, behavior design, creativity engine, governance, and compounding loops\" width=\"100%\">\n\u003C\u002Fp>\n\n**Third Brain V5 — 17 production-ready Agent Skills for Claude, Codex, Gemini, Cursor, and Windsurf. Build a persistent Knowledge OS with Obsidian, behavior design, creativity loops, verification, full-stack agent harnesses, and multi-agent workflows.**\n\nInstall 17 ready-to-use Agent Skills for ingesting sources, building an interlinked wiki, running daily review loops, verifying claims, managing context cost, improving agent workflows, and orchestrating agent teams.\n\n**Use this if your AI coding agent keeps forgetting context, making unverifiable \"done\" claims, or scattering useful knowledge across chats. Star it if you want a reusable, verification-first skill stack for long-running AI work.**\n\n[![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FMark393295827\u002Fthird-brain-v5-skills?style=social)](https:\u002F\u002Fgithub.com\u002FMark393295827\u002Fthird-brain-v5-skills\u002Fstargazers)\n[![Last commit](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flast-commit\u002FMark393295827\u002Fthird-brain-v5-skills)](https:\u002F\u002Fgithub.com\u002FMark393295827\u002Fthird-brain-v5-skills\u002Fcommits\u002Fmain)\n[![Issues](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fissues\u002FMark393295827\u002Fthird-brain-v5-skills)](https:\u002F\u002Fgithub.com\u002FMark393295827\u002Fthird-brain-v5-skills\u002Fissues)\n[![License: MIT](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-blue.svg)](LICENSE)\n[![Claude Code](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FClaude%20Code-Compatible-8A2BE2)](https:\u002F\u002Fclaude.ai\u002Fcode)\n[![Codex CLI](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FCodex%20CLI-Compatible-000000)](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fcodex)\n[![Gemini CLI](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGemini%20CLI-Compatible-4285F4)](https:\u002F\u002Fgithub.com\u002Fgoogle-gemini\u002Fgemini-cli)\n[![Cursor](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FCursor-Rules%20Adapter-111827)](docs\u002Fcompatibility.md)\n[![Windsurf](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FWindsurf-Skills%20Compatible-0EA5E9)](docs\u002Fcompatibility.md)\n[![Supported Models](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fmodels-Claude%20%7C%20GPT%20%7C%20Gemini-22C55E)](GUIDE.md)\n[![Agent Skills](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FAgent%20Skills-Format-2ea44f)](https:\u002F\u002Fagentskills.io)\n[![PRs Welcome](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPRs-welcome-brightgreen)](CONTRIBUTING.md)\n![Skills](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fskills-17-FF6B6B)\n![Karpathy LLM OS](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FKarpathy%20LLM%20OS-Compatible-8A2BE2)\n![Token Tracking](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Ftoken--tracking-00D1B2)\n![Ender's Game](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FEnder's%20Game-Command-8A2BE2)\n\n---\n\n## Visual Navigation\n\nOpen the local dashboard for a searchable skill map, workflow shortcuts, and cost tools:\n\n```bash\n# macOS\nopen tools\u002Findex.html\n\n# Windows PowerShell\nStart-Process tools\u002Findex.html\n```\n\nQuick links: [Visual Dashboard](tools\u002Findex.html) · [One-Click Test Prompt](examples\u002Fone-click-test-prompt.md) · [3-Minute Quickstart](examples\u002F3-minute-quickstart.md)\n\nFormat standard: [Agent Skills Standard](docs\u002Fagent-skills-standard.md) · [Agentic Engineering Principles](docs\u002Fagentic-engineering-principles.md) · run `python tools\u002Flint-agent-skills.py` before publishing skill changes.\n\n## Quick Install\n\n![Install demo](assets\u002Finstall-demo.svg)\n\nFor Claude Code, Codex CLI, or Gemini CLI:\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FMark393295827\u002Fthird-brain-v5-skills.git\ncd third-brain-v5-skills\nbash install.sh\n```\n\nExplicit targets:\n\n```bash\nbash install.sh codex\nbash install.sh claude\nbash install.sh gemini\nbash install.sh cursor\nbash install.sh windsurf\nbash install.sh all\n```\n\nManual install paths:\n\n```bash\n# Codex CLI\nmkdir -p ~\u002F.agents\u002Fskills && cp -r skills\u002F* ~\u002F.agents\u002Fskills\u002F\n\n# Claude Code\nmkdir -p ~\u002F.claude\u002Fskills && cp -r skills\u002F* ~\u002F.claude\u002Fskills\u002F\n\n# Gemini CLI\nmkdir -p ~\u002F.gemini\u002Fskills && cp -r skills\u002F* ~\u002F.gemini\u002Fskills\u002F\n\n# Cursor rules adapter\nmkdir -p .cursor\u002Frules && cp adapters\u002Fcursor\u002Fthird-brain-skills.mdc .cursor\u002Frules\u002F\n\n# Windsurf native workspace skills + routing rule\nmkdir -p .windsurf\u002Fskills .windsurf\u002Frules\ncp -r skills\u002F* .windsurf\u002Fskills\u002F\ncp adapters\u002Fwindsurf\u002Fthird-brain-skills.md .windsurf\u002Frules\u002F\n```\n\n## Try It\n\nPaste one of these into your agent after installing:\n\n```text\n\"I just read an interesting article. Ingest it into my wiki.\"\n\"Run my daily OKR.\"\n\"Cognitive compile on AI agent safety.\"\n\"Verify before I ship.\"\n\"Create an agent team to build a full-stack prototype.\"\n```\n\nOne-click test prompt: [examples\u002Fone-click-test-prompt.md](examples\u002Fone-click-test-prompt.md).\n\n## 3-Minute Quick Start\n\nThe fastest useful path is `wiki-ingest`: turn one article, PDF, transcript, or rough note into durable wiki knowledge. Treat this as the recommended first skill; add the rest only after the previous layer is producing evidence.\n\n```text\nUse wiki-ingest on this source.\n\nGoal: turn it into reusable wiki knowledge, not a loose summary.\n\nCreate:\n1. one immutable source note in `SOURCES_DIR` (default: `sources\u002F`)\n2. 3-7 key insights with source references\n3. at least one concept page in `CONCEPTS_DIR` (default: `wiki\u002Fconcepts\u002F`)\n4. relevant entity pages in `ENTITIES_DIR` (default: `wiki\u002Fentities\u002F`)\n5. links from the new pages to existing related pages when possible\n6. a short log entry in `LOG_FILE` (default: `system\u002Flog.md`)\n\nAfter writing files, run a quick verification:\n- list created or updated files\n- check each wiki page has at least two wikilinks\n- state any claims that are single-source only\n```\n\nFull copyable workflow: [examples\u002F3-minute-quickstart.md](examples\u002F3-minute-quickstart.md).\n\n## Skill Adoption Ladder\n\n| Stage | Core skills | Upgrade when |\n|---|---|---|\n| Week 1: Start | `wiki-ingest` + `verify-before-claim` | You can ingest at least one source per day and every completion claim has fresh evidence. |\n| Weeks 2-4: Daily loop | + `daily-okr` + `session-learn` | Daily OKR score is above 70% for a week and session learnings are being written back. |\n| Month 2+: Deepen | + `cognitive-compile` + `behavior-design` + `creativity-engine` | The wiki has 50+ pages or repeated decisions need deeper synthesis and experiments. |\n| Month 3+: Engineer | + `knowledge-ops` + `harness-engineering` + `agentic-engineering` | Retrieval, permissions, delegated actions, provenance, or workflow reliability become bottlenecks. |\n| Multi-agent: Scale | + `agent-teams-command` + `project-flow-ops` | Work can be split into separate owners with clear integration and verification gates. |\n| Strategy: Evaluate | + `startup-evaluation` + `anthropic-os` + `deep-research` | You need startup health, market, operating-system, or STOW-compatible evidence research decisions. |\n\n## Before → After\n\n| Use Case | Before | After |\n|----------|--------|-------|\n| Research PDF | A long PDF is summarized once and then forgotten. | `wiki-ingest` creates source notes, concept pages, entity pages, links, and a reusable brief. |\n| Coding session | The agent says \"fixed\" without proof. | `verify-before-claim` requires a fresh command, exit code, and evidence before any completion claim. |\n| Daily knowledge work | Inputs, ideas, and actions stay scattered across chat history. | `daily-okr` produces one insight, one wiki update, one action, one reusable output, and a daily score. |\n\n## Demo & Tools\n\n| Resource | What it shows |\n|----------|---------------|\n| [tools\u002Findex.html](tools\u002Findex.html) | Visual dashboard and skill navigation |\n| [dashboard.html](dashboard.html) | Visual overview of the Third Brain V5 system |\n| [tools\u002Ftoken-calculator.html](tools\u002Ftoken-calculator.html) | Token cost calculator |\n| [tools\u002Fpipeline-cost-calculator.html](tools\u002Fpipeline-cost-calculator.html) | Pipeline cost estimator |\n| [GUIDE.md](GUIDE.md) | Full installation, workflow, and troubleshooting guide |\n| [examples\u002F](examples\u002F) | Copyable workflows and Obsidian entry note |\n| [docs\u002Fcompatibility.md](docs\u002Fcompatibility.md) | Cursor, Windsurf, and other AI IDE setup |\n| [docs\u002Fcommunity-discovery.md](docs\u002Fcommunity-discovery.md) | GitHub description, topics, and Awesome-list submission notes |\n| [docs\u002Fgithub-star-growth-sprint.md](docs\u002Fgithub-star-growth-sprint.md) | 48-hour launch and Awesome-list growth sprint |\n\nIf this saves you time, consider starring the repo so others can discover it through GitHub search.\n\n---\n\n## Overview\n\nThese skills transform any AI coding agent into a **persistent knowledge compounding system**. Instead of treating each conversation as a one-off chat, these skills create an interlinked knowledge base that grows richer with every session — while also providing strategic execution frameworks inspired by Anthropic's work methods and Karpathy's Agentic Engineering.\n\n### Agent Understanding Framework\n\nThird Brain treats agents as LLM OS processes, not chat personas: the LLM is the CPU, context is RAM, Obsidian\u002Fwiki\u002Flogs are disk, tools are system calls, skills are executable programs, and the harness is the kernel that enforces permissions and observability. See [Agent Understanding Framework](wiki\u002Fconcepts\u002Fagent-understanding-framework.md).\n\nThe latest engineering update adds a full-stack agent lens: agent workflows now account for IDE subagents, personal agents, agentic search, commerce mandates, generative-media provenance, and ambient-device privacy. See [Agentic Engineering Principles](docs\u002Fagentic-engineering-principles.md).\n\n### Design Philosophy\n\n| Layer | Principle | Skills |\n|-------|-----------|--------|\n| **🧠 Knowledge** | Capture, structure, and compound knowledge over time | wiki-ingest, knowledge-ops, wiki-lint |\n| **⚡ Daily Loop** | Close the knowledge-to-action cycle every day | daily-okr, cognitive-compile |\n| **🎯 Behavior & Creativity** | Turn knowledge into habits and novel ideas | behavior-design, creativity-engine |\n| **🔬 Research & Quality** | Verify before claiming, research with rigor | deep-research, verify-before-claim |\n| **🔄 Continuous Learning** | Extract patterns from every session | session-learn, project-flow-ops |\n| **📊 Context & Cost** | Manage the LLM's scarcest resource | context-manager, token-cost-tracker |\n| **🏗️ Engineering** | Design agent workflows, infrastructure, and multi-agent teams | agentic-engineering, harness-engineering, agent-teams-command |\n| **💼 Strategy** | Evaluate startups, adopt Anthropic-level methods | startup-evaluation, anthropic-os |\n\n---\n\n## Skills\n\n### 📥 Ingestion & Knowledge\n\n| Skill | Description |\n|-------|-------------|\n| [wiki-ingest](skills\u002Fwiki-ingest\u002FSKILL.md) | Ingest sources into a persistent interlinked wiki with source-risk taxonomy, macro-action scope, concept\u002Fentity pages, Karpathy understanding gate, clipping archive, and post-ingest lint. |\n| [knowledge-ops](skills\u002Fknowledge-ops\u002FSKILL.md) | Multi-layer knowledge management — classify, deduplicate, preserve evidence hierarchy, keep Markdown-first retrieval primary, use vectors as optional support, and queue knowledge debt. |\n| [wiki-lint](skills\u002Fwiki-lint\u002FSKILL.md) | Health-check the wiki across P0\u002FP1 graph health, frontmatter, source refs, links, clipping lifecycle, provenance debt, understanding integrity, contradictions, and drift. |\n\n### 🔄 Daily Workflow\n\n| Skill | Description |\n|-------|-------------|\n| [daily-okr](skills\u002Fdaily-okr\u002FSKILL.md) | Daily knowledge compound closed loop — 7 Key Results from input to feedback, with scoring. Includes Stop Doing List (Buffett\u002FMunger). |\n| [cognitive-compile](skills\u002Fcognitive-compile\u002FSKILL.md) | Deep learning 8-section framework: Question → Facts → Concepts → Pattern Recognition → Conflict Detection → Hypothesis Generation → Decision Support → Action. |\n\n### 🎨 Behavior & Creativity\n\n| Skill | Description |\n|-------|-------------|\n| [behavior-design](skills\u002Fbehavior-design\u002FSKILL.md) | Transform goals into behavior systems. Decompose → minimum habits → triggers → SOPs → review. Includes HAS (Human Agency Scale) from Stanford research. |\n| [creativity-engine](skills\u002Fcreativity-engine\u002FSKILL.md) | Generate novel ideas via combinatorial creativity. Lego Building Blocks method (Andrew Ng), cross-domain analogies, minimum experiments. |\n\n### 🔬 Research & Quality\n\n| Skill | Description |\n|-------|-------------|\n| [deep-research](skills\u002Fdeep-research\u002FSKILL.md) | STOW-compatible research harness with ChatGPT-style preflight, source\u002Fclaim ledgers, activity trace, citations, privacy checks, and wiki-ingest handoff packets. |\n| [verify-before-claim](skills\u002Fverify-before-claim\u002FSKILL.md) | Iron rule: No completion claims without fresh verification evidence. Includes expected value thinking from poker psychology. |\n\n### 🔄 Learning & Flow\n\n| Skill | Description |\n|-------|-------------|\n| [session-learn](skills\u002Fsession-learn\u002FSKILL.md) | Extract knowledge patterns from sessions — 7 signal types (concepts, entities, corrections, patterns, ideas, decisions, gaps). Closure Protocol included. |\n| [project-flow-ops](skills\u002Fproject-flow-ops\u002FSKILL.md) | Execution flow — triage, plan, track, review across projects. |\n\n### 📊 Context & Cost\n\n| Skill | Description |\n|-------|-------------|\n| [context-manager](skills\u002Fcontext-manager\u002FSKILL.md) | Manage the LLM's context window — token budgeting, prompt assembly, truncation strategies. Concrete Ideas framework (Andrew Ng) + Tokenmaxxing vs Efficiency (Gary Tan). |\n| [token-cost-tracker](commands\u002Ftoken-cost-tracker.md) | Estimate, log, and report token usage. Built-in Python logger script. |\n\n### 🏗️ Engineering\n\n| Skill | Description |\n|-------|-------------|\n| [agentic-engineering](skills\u002Fagentic-engineering\u002FSKILL.md) | Refactor workflows into spec-driven macro actions with quality ceilings, delegated-action boundaries, verification gates, state checkpoints, and write-back. |\n| [harness-engineering](skills\u002Fharness-engineering\u002FSKILL.md) | Design runtime infrastructure around AI agents — permissions, system-call tools, delegated-action gates, provenance, observability, recovery, and adversarial review. |\n| [agent-teams-command](skills\u002Fagent-teams-command\u002FSKILL.md) | Command multi-agent macro actions with ownership, IPC, async budget envelopes, integration joins, cleanup, evidence gates, and red-team review for high-risk work. |\n\n### 💼 Strategy\n\n| Skill | Description |\n|-------|-------------|\n| [startup-evaluation](skills\u002Fstartup-evaluation\u002FSKILL.md) | Evaluate startup health with entrepreneurship, VC 5T, PMF, runway, team, unit economics, and next-cheapest-test diagnostics. |\n| [anthropic-os](skills\u002Fanthropic-os\u002FSKILL.md) | Anthropic OS — Self-Evolving Work Method Engine. CASH growth system, 70\u002F30 rule, two-week rule, hive mind, working backwards. Built-in self-evolution mechanism. |\n\n---\n\n## Architecture\n\n```\n┌──────────────────────────────────────────────────────────────────────────┐\n│                      Karpathy LLM OS Layer                                │\n│  LLM=CPU │ Context=RAM │ Storage=Disk │ Tools=System Calls                │\n│  Skills=Programs │ Harness=Kernel │ Agent Teams=Processes                 │\n│  ┌──────────────────────────────────────────────────────────────────┐    │\n│  │ context-manager: Token Budget → Prompt Assembly → Truncation     │    │\n│  │ token-cost-tracker: Estimate → Log → Report                     │    │\n│  └──────────────────────────────────────────────────────────────────┘    │\n└──────────────────────────────────────────────────────────────────────────┘\n                                    │\n                         ┌──────────┴──────────┐\n                         ▼                     ▼\n              ┌──────────────────┐   ┌──────────────────────┐\n              │   External       │   │   Agent Teams        │\n              │   Sources        │   │   (Parallel Fleet)    │\n              └────────┬─────────┘   └──────────────────────┘\n                       ▼\n         ┌──────────────────────────────┐\n         │   wiki-ingest + knowledge-ops│\n         │   (STOW pipeline + RAG sync) │\n         └──────┬──────────┬────────────┘\n                │          │\n         ┌──────▼          └──────────────┐\n         │  Knowledge Layers               │\n         │  ├ Active (GitHub\u002FLinear)       │\n         │  ├ Memory (quick access)        │\n         │  ├ Wiki (durable, interlinked)  │\n         │  ├ Vector (ChromaDB, semantic)  │\n         │  └ External (DBs, APIs)         │\n         └────────────────────────────────┘\n                │\n    ┌───────────┼──────────┬──────────────┬──────────────┐\n    ▼           ▼          ▼              ▼              ▼\n┌─────────┐ ┌─────────┐ ┌──────────┐ ┌───────────┐ ┌──────────┐\n│ daily   │ │cognitive│ │ behavior │ │ creativity│ │ project  │\n│ -okr    │ │-compile │ │ -design  │ │ -engine   │ │ -flow-ops│\n└─────────┘ └─────────┘ └──────────┘ └───────────┘ └──────────┘\n    │           │          │              │              │\n    └───────────┼──────────┼──────────────┼──────────────┘\n                ▼\n    ┌─────────────────────────────────────────────────────────────┐\n    │ session-learn (+Closure Protocol)  ← feedback loop          │\n    │ verify-before-claim               ← quality gate            │\n    │ wiki-lint                         ← health check            │\n    │ deep-research                     ← research + STOW handoff │\n    │ harness-engineering               ← safety + multi-agent    │\n    │ agent-teams-command               ← fleet command           │\n    │ startup-evaluation                ← VC evaluation           │\n    │ anthropic-os                      ← work method engine      │\n    └─────────────────────────────────────────────────────────────┘\n```\n\n---\n\n## Quick Start\n\nFor detailed installation, usage, and 5 complete workflow scenarios, see the **[Full Guide](GUIDE.md)**.\n\n### Clone\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FMark393295827\u002Fthird-brain-v5-skills.git\ncd third-brain-v5-skills\nbash install.sh\n```\n\n### Claude Code\n\n```bash\n# Personal skills (available across all projects)\ncp -r third-brain-v5-skills\u002Fskills\u002F* ~\u002F.claude\u002Fskills\u002F\n\n# Project skills (shared with team)\ncp -r third-brain-v5-skills\u002Fskills\u002F* .claude\u002Fskills\u002F\n```\n\n### Codex CLI\n\n```bash\ncp -r third-brain-v5-skills\u002Fskills\u002F* ~\u002F.agents\u002Fskills\u002F\n```\n\n### Gemini CLI\n\n```bash\ncp -r third-brain-v5-skills\u002Fskills\u002F* ~\u002F.gemini\u002Fskills\u002F\n```\n\n### Try It\n\n```\n\"I just read an interesting article — ingest it into my wiki.\"\n\"Run my daily OKR.\"\n\"Cognitive compile on AI Agent safety.\"\n\"Design a reading habit.\"\n\"Generate 10 business ideas from my wiki.\"\n\"Extract what we learned this session.\"\n\"Verify before I ship.\"\n\"Lint my wiki.\"\n\"Create an agent team to build a full-stack app.\"\n\"Launch Anthropic OS for my team's growth.\"\n```\n\n---\n\n## Command Reference\n\n| Command | File | Usage |\n|---------|------|-------|\n| `token-cost-tracker estimate` | [token-cost-tracker.md](commands\u002Ftoken-cost-tracker.md) | Estimate token cost before a task |\n| `token-cost-tracker log` | [token-cost-tracker.md](commands\u002Ftoken-cost-tracker.md) | Log actual token usage |\n| `token-cost-tracker report` | [token-cost-tracker.md](commands\u002Ftoken-cost-tracker.md) | Weekly\u002Fmonthly cost report |\n\n---\n\n## Wiki Structure\n\nThese skills use `system\u002Fconfig.md` as the default path contract. The layout below is the default, not a hard requirement; override it in your local `system\u002Fconfig.md`, `CLAUDE.md`, or equivalent agent rules file if your Obsidian vault already uses different folders.\n\n```\nsources\u002F          ← Immutable source notes (articles, books, reports)\nwiki\u002F\n├── concepts\u002F     ← Ideas, frameworks, methods\n├── entities\u002F     ← People, companies, products\n├── atomic-notes\u002F ← Single knowledge atoms\n├── outputs\u002F      ← Reusable outputs (reports, analyses, evaluations)\n├── decisions\u002F    ← Architecture and strategy decisions\n└── sops\u002F         ← Standard operating procedures\nmaps\u002F             ← Navigation \u002F Maps of Content\nsystem\u002F           ← Schema, log, templates, lint reports\n08_behaviors\u002F     ← Behavior system (goals, habits, SOPs, reviews)\n09_creativity\u002F    ← Creativity system (ideas, experiments, prototypes)\n```\n\nPath config: [system\u002Fconfig.md](system\u002Fconfig.md).\n\n---\n\n## Related Projects\n\n- [Agent Skills](https:\u002F\u002Fagentskills.io) — Open format specification\n- [llm-wiki-agent](https:\u002F\u002Fgithub.com\u002FSamurAIGPT\u002Fllm-wiki-agent) — Original STOW pattern implementation\n\n## Contributing\n\nBug reports, skill requests, and small PRs are welcome. See [CONTRIBUTING.md](CONTRIBUTING.md), [CHANGELOG.md](CHANGELOG.md), and the [release playbook](docs\u002Frelease-playbook.md) for the weekly feedback loop and release checklist.\n\n## License\n\nMIT — see [LICENSE](LICENSE).\n","Third Brain V5 Skills 是一个提供17种即用型AI代理技能的项目，旨在帮助用户构建持久的知识操作系统。其核心功能包括源内容摄入、构建互链式wiki、执行日常审查循环、验证声明、管理上下文成本、改进代理工作流以及协调多代理团队等。技术特点在于支持Claude, Codex, Gemini, Cursor, 和Windsurf等多种AI模型，并通过Obsidian进行知识管理和行为设计。适用于需要长期运行AI任务且希望保持上下文一致性、避免知识碎片化散落于聊天记录中的场景。","2026-06-11 04:08:48","CREATED_QUERY"]