{"id":50669798,"url":"https://github.com/kbwen/agentic-os","last_synced_at":"2026-06-08T10:01:08.962Z","repository":{"id":350787547,"uuid":"1208266947","full_name":"KbWen/agentic-os","owner":"KbWen","description":"Governance-first OS for AI coding agents — structured workflows, delivery gates, engineering guardrails, and 17 professional skills for Claude Code, Cursor, Copilot, Antigravity \u0026 Codex.","archived":false,"fork":false,"pushed_at":"2026-06-02T04:29:54.000Z","size":1182,"stargazers_count":3,"open_issues_count":31,"forks_count":2,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-06-02T05:12:34.161Z","etag":null,"topics":["agent-framework","agentic-development","agents-md","ai-agent","ai-governance","ai-guardrails","ai-workflow","claude-code","coding-agent","cursor-rules","developer-tools","google-antigravity","llm-tools","multi-agent","token-optimization"],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/KbWen.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":"CODE_OF_CONDUCT.md","threat_model":null,"audit":null,"citation":"CITATION.cff","codeowners":".github/CODEOWNERS","security":"SECURITY.md","support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":"AGENTS.md","dco":null,"cla":null}},"created_at":"2026-04-12T03:32:25.000Z","updated_at":"2026-06-02T04:29:57.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/KbWen/agentic-os","commit_stats":null,"previous_names":["kbwen/agentic-os"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/KbWen/agentic-os","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KbWen%2Fagentic-os","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KbWen%2Fagentic-os/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KbWen%2Fagentic-os/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KbWen%2Fagentic-os/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/KbWen","download_url":"https://codeload.github.com/KbWen/agentic-os/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KbWen%2Fagentic-os/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34057158,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-08T02:00:07.615Z","response_time":111,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agent-framework","agentic-development","agents-md","ai-agent","ai-governance","ai-guardrails","ai-workflow","claude-code","coding-agent","cursor-rules","developer-tools","google-antigravity","llm-tools","multi-agent","token-optimization"],"created_at":"2026-06-08T10:00:29.741Z","updated_at":"2026-06-08T10:01:08.956Z","avatar_url":"https://github.com/KbWen.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003cp align=\"center\"\u003e\n  \u003cimg src=\"https://img.shields.io/badge/Agentic%20OS-v1.4.0-blueviolet?style=for-the-badge\u0026logo=data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHdpZHRoPSIyNCIgaGVpZ2h0PSIyNCIgdmlld0JveD0iMCAwIDI0IDI0IiBmaWxsPSJ3aGl0ZSI+PHBhdGggZD0iTTEyIDJDNi40OCAyIDIgNi40OCAyIDEyczQuNDggMTAgMTAgMTAgMTAtNC40OCAxMC0xMFMxNy41MiAyIDEyIDJ6bTAgMThjLTQuNDEgMC04LTMuNTktOC04czMuNTktOCA4LTggOCAzLjU5IDggOC0zLjU5IDgtOCA4eiIvPjwvc3ZnPg==\" alt=\"Agentic OS v1.4.0\"/\u003e\n\u003c/p\u003e\n\n\u003ch1 align=\"center\"\u003eAgentic OS\u003c/h1\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003cstrong\u003eA governance-first layer for AI coding agents.\u003c/strong\u003e\u003cbr/\u003e\n  Portable workflows, delivery gates, engineering guardrails, and 14 professional skills\u003cbr/\u003e\n  for Claude Code, OpenAI Codex, Google Antigravity, Cursor, GitHub Copilot, and other coding agents.\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://github.com/KbWen/agentic-os/actions/workflows/validate.yml\"\u003e\u003cimg src=\"https://img.shields.io/github/actions/workflow/status/KbWen/agentic-os/validate.yml?branch=main\u0026style=flat-square\u0026label=CI\" alt=\"CI Status\"/\u003e\u003c/a\u003e\n  \u003ca href=\"https://github.com/KbWen/agentic-os/actions/workflows/security.yml\"\u003e\u003cimg src=\"https://img.shields.io/github/actions/workflow/status/KbWen/agentic-os/security.yml?branch=main\u0026style=flat-square\u0026label=Security\" alt=\"Security Scan\"/\u003e\u003c/a\u003e\n  \u003ca href=\"https://opensource.org/licenses/MIT\"\u003e\u003cimg src=\"https://img.shields.io/badge/License-MIT-22c55e?style=flat-square\" alt=\"MIT License\"/\u003e\u003c/a\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"docs/README_zh-TW.md\"\u003e繁體中文\u003c/a\u003e \u0026middot;\n  \u003ca href=\"https://github.com/KbWen/agentic-os/blob/main/CONTRIBUTING.md\"\u003eContributing\u003c/a\u003e \u0026middot;\n  \u003ca href=\"https://github.com/KbWen/agentic-os/blob/main/CHANGELOG.md\"\u003eChangelog\u003c/a\u003e\n\u003c/p\u003e\n\n---\n\n## The Problem\n\nAI coding agents are powerful, but they need shared operating rules. Without structure, they can:\n\n- **Skip steps** — jump straight to code without planning or reviewing\n- **Hallucinate completion** — claim \"done\" without verifiable evidence\n- **Drift from scope** — refactor code nobody asked them to touch\n- **Lose context** — forget decisions across conversations, forcing re-derivation\n- **Break things silently** — no safety gates between \"idea\" and \"production\"\n\n## The Solution\n\n**Agentic OS** is a portable governance framework that gives AI agents a repeatable engineering workflow. Install it into your project, and your AI agents gain:\n\n```mermaid\nflowchart LR\n    U[\"User\u003cbr/\u003esays\"] --\u003e G{\"Gate\u003cbr/\u003eEngine\"}\n    G --\u003e|PASS| W[\"Workflow\u003cbr/\u003e+ Skills\"]\n    W --\u003e E{\"Evidence\u003cbr/\u003eRequired\"}\n    E --\u003e|PASS| S[\"Ship\u003cbr/\u003e→ SSoT\"]\n    G --\u003e|FAIL| X1[\"⛔ STOP\"]\n    E --\u003e|FAIL| X2[\"⛔ STOP\"]\n\n    style U fill:#8b5cf6,color:#fff,stroke:none\n    style G fill:#f59e0b,color:#fff,stroke:none\n    style W fill:#3b82f6,color:#fff,stroke:none\n    style E fill:#22c55e,color:#fff,stroke:none\n    style S fill:#06b6d4,color:#fff,stroke:none\n    style X1 fill:#ef4444,color:#fff,stroke:none\n    style X2 fill:#ef4444,color:#fff,stroke:none\n```\n\nThe idea behind every phase is the same: if there's no verifiable evidence, the task isn't done — and the gates enforce that, instead of trusting the agent to self-report honestly.\n\n---\n\n## Features\n\n### Gate Engine \u0026 Phase System\n\nEvery task flows through mandatory phases. The AI cannot skip ahead.\n\n```mermaid\nflowchart LR\n    B[\"/bootstrap\"] --\u003e P[\"/plan\"]\n    P --\u003e I[\"/implement\"]\n    I --\u003e R[\"/review\"]\n    R --\u003e T[\"/test\"]\n    T --\u003e S[\"/ship\"]\n\n    style B fill:#8b5cf6,color:#fff,stroke:none\n    style P fill:#3b82f6,color:#fff,stroke:none\n    style I fill:#22c55e,color:#fff,stroke:none\n    style R fill:#f59e0b,color:#fff,stroke:none\n    style T fill:#ef4444,color:#fff,stroke:none\n    style S fill:#06b6d4,color:#fff,stroke:none\n```\n\n| Classification | Required Phases |\n|:---|:---|\n| **tiny-fix** | Classify → Execute → Evidence → Done |\n| **quick-win** | Bootstrap → Plan → Implement → Evidence → Ship |\n| **feature** | Bootstrap → Spec → Plan → Implement → Review → Test → Handoff → Ship |\n| **hotfix** | Bootstrap → Research → Plan → Implement → Review → Test → Ship |\n| **architecture-change** | Bootstrap → ADR → Spec → Plan → Implement → Review → Test → Handoff → Ship |\n\n### Engineering Guardrails\n\nA constitution for AI behavior — loaded automatically, enforced at every phase:\n\n- **No Evidence = No Completion** — narrative claims are not proof\n- **Scope Discipline** — unauthorized refactoring is strictly prohibited\n- **Destructive Command Blocking** — `rm -rf`, `git reset --hard`, force pushes require pre-approved rollback plans\n- **OWASP Top 10 Auto-Scan** — security checks run during `/implement` and `/review`\n- **Confidence Gate** — AI must declare confidence level; low confidence triggers escalation\n\n### 14 Professional Skills\n\nSkills auto-activate based on task classification and workflow phase:\n\n| Skill | Trigger | Description |\n|:---|:---|:---|\n| Test-Driven Development | feature, architecture-change | Red → Green → Refactor cycles |\n| Systematic Debugging | bug encounter | 4-phase root cause analysis |\n| Red Team / Adversarial | review, test | Classification-based security analysis |\n| API Design | API endpoints detected | Endpoint validation enforcement |\n| Auth Security | auth code detected | Hashing, tokens, rate limiting |\n| Database Design | migration detected | Forward-only ORM-aware migration safety |\n| Frontend Patterns | UI components | Component and state management patterns |\n| Parallel Agent Dispatching | complex tasks | Coordinated subagent execution |\n| Subagent-Driven Development | multi-module tasks | Multi-agent coordination |\n| Karpathy Principles | all coding tasks | Behavioral guardrails against common LLM coding mistakes |\n| Production Readiness | feature, architecture-change | Pre-ship observability: error sinks, log strategy, rollback telemetry |\n| Verification Before Completion | /ship | 5-gate check: Scope → Quality → Evidence → Risk → Communication |\n| Git Worktrees | parallel branches | Worktree isolation workflows |\n| Doc Lookup | documentation needed | Documentation retrieval strategy |\n\n### Single Source of Truth (SSoT)\n\nEvery project has one canonical state file. AI agents read it first, write to it last.\n\n```\n.agentcortex/context/\n├── current_state.md          # Global project state (SSoT)\n└── work/\n    └── \u003cbranch-name\u003e.md      # Per-task work log (isolated)\n```\n\n- **Work Logs** track per-task progress, evidence, and gate receipts\n- **SSoT** tracks global decisions, lessons, and ship history\n- **Handoff** enables seamless AI-to-AI continuity across conversations\n\n### Multi-Agent Collaboration\n\nBuilt for teams where multiple AI sessions work on the same codebase:\n\n- **One Branch = One Owner** — prevents concurrent Work Log corruption\n- **Advisory Locking** — lock files signal active sessions without blocking\n- **Ship Guard** — checks for SSoT conflicts before merging\n- **Session Identity** — every AI session writes its model name and timestamp\n\n### Three Entry Paths\n\nPick the one that matches your starting point (full opener templates in **Quick Start §2**):\n\n| Starting point | First command | Full chain |\n|---|---|---|\n| 🆕 Brand-new project, multi-feature idea | `/spec-intake` | spec-intake → pick feature → bootstrap → app-init → plan → implement |\n| 🏗️ Existing repo, adopting Agentic OS | `/audit` (read-only, zero risk) | audit → app-init → spec-intake → quick-win → feature |\n| 🎯 Single concrete task on an established project | `/bootstrap` | bootstrap → plan → implement → review → test → ship |\n\nSee the [Lifecycle Benchmark](https://github.com/KbWen/agentic-os/blob/main/docs/LIFECYCLE_BENCHMARK.md) ([繁體中文](https://github.com/KbWen/agentic-os/blob/main/docs/LIFECYCLE_BENCHMARK_zh-TW.md)) for real token consumption data across 6 development scenarios.\n\n### Token Efficiency\n\nDesigned for cost-effective models (Gemini Flash, Haiku, etc.):\n\n- **Conditional Loading** — tiny-fix skips guardrails (~5,000 tokens saved)\n- **Skill Cache Policy** — metadata-first loading, full SKILL.md only on cache miss\n- **Phase Summary** — compact 1-liner per phase for low-token resume\n- **Read-Once Discipline** — governance docs persist in context, never re-read\n\n---\n\n## Quick Start\n\n### 1. Install\n\n**Prerequisites**:\n\n| Dependency | Required? | Purpose |\n|:---|:---|:---|\n| **Git** | Required | Clone and deploy the framework |\n| **Bash** | Required | Run deploy \u0026 validate scripts (Git Bash from [Git for Windows](https://gitforwindows.org/) is enough on Windows) |\n| **Python 3.9+** | Recommended | Enables full validation (metadata, encoding, command sync checks) |\n| **SHA-256 tool** | Required for deploy | `sha256sum`, `shasum`, or `openssl` (pre-installed on most systems) |\n\n\u003e **No Python?** The framework deploys and works without Python. Validation runs in\n\u003e degraded mode — Python-dependent checks report `WARN` instead of `FAIL`.\n\u003e Pass `--no-python` to suppress warnings: `bash .agentcortex/bin/validate.sh --no-python`\n\n**Install (first time):**\n\n```bash\n# Clone Agentic OS\ngit clone https://github.com/KbWen/agentic-os.git\n\n# Preview what will be deployed (no changes made)\n./agentic-os/installers/deploy_brain.sh --dry-run /path/to/your-project\n\n# Deploy into your project\n./agentic-os/installers/deploy_brain.sh /path/to/your-project\n```\n\n**Update (after first install):** the installer lives inside your project. Run it from your project root — it reads the deploy manifest and auto-fetches the latest framework version from GitHub.\n\n```bash\nbash installers/deploy_brain.sh .\n```\n\n\u003e **Existing files won't be overwritten.** If your project already has `AGENTS.md`, `CLAUDE.md`, or other framework-managed files, they are preserved. The new framework version is saved as `\u003cfilename\u003e.acx-incoming` sidecar. Review and merge manually — or ask your AI agent: *\"Merge each .acx-incoming into its target, preserving my project-specific content and adopting framework updates.\"*\n\n\u003e **AI-agent install:** If you're asking an AI assistant to install Agentic OS, point it to this README. The commands above are deterministic — no platform-specific heuristics required.\n\n**Optional local pre-commit validation:** enable the bundled Git hook sample to run Agentic OS validation before each commit.\n\n```bash\ncp .githooks/pre-commit.guard-ssot.sample .githooks/pre-commit\nchmod +x .githooks/pre-commit\ngit config core.hooksPath .githooks\n```\n\nOn Windows, run the setup from PowerShell:\n\n```powershell\nCopy-Item .githooks\\pre-commit.guard-ssot.sample .githooks\\pre-commit\ngit config core.hooksPath .githooks\n```\n\nThe hook runs `validate.ps1` from Git Bash on Windows when PowerShell is available, otherwise it runs `validate.sh`. Validator failures block the commit; guarded SSoT receipt warnings remain advisory.\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cb\u003eWindows (PowerShell / CMD)\u003c/b\u003e\u003c/summary\u003e\n\n```powershell\n# Clone Agentic OS (first time)\ngit clone https://github.com/KbWen/agentic-os.git\n\n# Preview what will be deployed (no changes made)\npowershell -ExecutionPolicy Bypass -File .\\agentic-os\\installers\\deploy_brain.ps1 -DryRun C:\\path\\to\\your-project\n\n# Deploy into your project\npowershell -ExecutionPolicy Bypass -File .\\agentic-os\\installers\\deploy_brain.ps1 C:\\path\\to\\your-project\n\n# CMD alternative\n.\\agentic-os\\installers\\deploy_brain.cmd C:\\path\\to\\your-project\n```\n\nUse the PowerShell entrypoint when possible. It resolves Git Bash directly and does not require a WSL distro.\nGit Bash is still required on Windows for shell-based deploy and validation scripts; the PowerShell entrypoint simply locates and invokes it for you.\n\n```powershell\n# Already installed? Run from your project root to update:\npowershell -ExecutionPolicy Bypass -File .\\installers\\deploy_brain.ps1 .\n\n# Validation after install\npowershell -ExecutionPolicy Bypass -File .\\.agentcortex\\bin\\validate.ps1\n\n# Lightweight validation when Python is not installed\nbash ./.agentcortex/bin/validate.sh --no-python\n```\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cb\u003eText-only usage (no scripts)\u003c/b\u003e\u003c/summary\u003e\n\nIf you only want the governance templates (Markdown files) without running any tooling:\n\n1. Copy the `.agent/`, `.agents/`, and `AGENTS.md` files into your project\n2. Optionally copy `.agentcortex/context/` and `.agentcortex/templates/` for state management\n3. No Python, Bash, or other tools are needed — all governance is plain Markdown\n\n\u003c/details\u003e\n\n### 1a. Customizing Without Conflicts (Fork or Clone)\n\nHowever you adopt Agentic OS — **fork** the repo, or **clone + `deploy_brain.sh`** into your project — the same rule keeps upgrades painless: **add your own files; never edit framework-owned files in place.** Put your customizations where the framework guarantees never to touch them, and they survive both `git pull upstream` (fork) and the next `deploy` (clone):\n\n| You want to… | Put it here | Why it survives upgrades |\n|---|---|---|\n| Add project governance (narrow/disable a directive) | `AGENTS.override.md` (project root) or `~/.agentcortex/AGENTS.override.md` (personal) | Loaded present-only at session start; framework never ships these files. MAY narrow/disable directives but **cannot** relax delivery gates. |\n| Add your own skills | `.agents/skills/custom-\u003cname\u003e/SKILL.md` (+ `.agent/skills/custom-\u003cname\u003e` metadata) | `custom-*` is a reserved namespace the framework never ships → zero collision, never overwritten. |\n| Adjust skill activation (pin/exclude) | `.agentcortex/context/private/user-preferences.yaml` | Gitignored, personal, loaded by bootstrap. |\n\n**What NOT to do:** editing a framework file in place (`AGENTS.md`, `.agent/rules/*`, `.agent/workflows/*`, a shipped skill body) causes merge conflicts on `git pull upstream` (fork) and is force-updated on the next `deploy` (clone). Framework **skills** are the one tolerant exception — if you edit one, `deploy` preserves your copy as a visible `\u003cfile\u003e.acx-incoming` sidecar rather than silently overwriting it — but the cleaner pattern is always to copy it to a `custom-\u003cname\u003e` skill and edit that. Governance, security, and workflow files always force-update so you keep receiving fixes.\n\n### 2. Start Working — Pick Your Entry Point\n\nThe right first command depends on your starting point. **Always preface with**:\n\n\u003e \"Read `AGENTS.md` and follow it. Do not claim completion until /review and /test pass.\"\n\nThen add **one** of the following based on your situation:\n\n**A. Brand-new project, multi-feature idea** — `/spec-intake` first\n\n```text\nThis is a brand-new project. My initial idea is:\n[1–2 paragraphs describing the product and its features]\n\nPlease run /spec-intake to decompose this into a feature inventory.\nAfter I pick the first feature, run /bootstrap → /app-init to establish\nthe tech stack ADR, then /plan and /implement.\n```\n\n**B. Existing repository adopting Agentic OS** — `/audit` first\n\n```text\nThis repo already has code; Agentic OS was just deployed.\nRun /audit (read-only) to map the codebase, then /app-init to record\nthe tech stack ADR, then /spec-intake when I add features.\n```\n\n**C. Single concrete task on an established project** — `/bootstrap` directly\n\n```text\n/bootstrap\n[describe the single task]\n```\n\n\u003e **Why this matters**: a multi-feature raw idea routed straight to `/bootstrap`\n\u003e is classified as a single task, skipping the Feature Inventory and\n\u003e `_product-backlog.md` decomposition. The Intent Router auto-detects this in\n\u003e most cases, but explicit beats inferred.\n\n### 3. Follow the Flow\n\nThe AI will classify your task and follow the required phases automatically:\n\n```\nYou: \"Add pagination to the user list API\"\n\nAI: [/bootstrap] → Classification: feature\n    → Required: Bootstrap → Spec → Plan → Implement → Review → Test → Handoff → Ship\n    → Loading skills: API Design, Test-Driven Development\n    → ⚡ ACX\n```\n\n---\n\n## Commands\n\n| Command | Purpose |\n|:---|:---|\n| `/bootstrap` | Initialize task, classify scope, create work log |\n| `/spec-intake` | Import and decompose external specs |\n| `/spec` | Define verifiable specifications |\n| `/plan` | Create implementation plan with rollback steps |\n| `/implement` | Execute code changes (gate-protected) |\n| `/review` | Logic, security, and scope audit |\n| `/test` | Verify with minimal necessary tests |\n| `/handoff` | Resumable state summary for next session |\n| `/ship` | Consolidate evidence, update SSoT, merge |\n| `/adr` | Architecture Decision Record |\n| `/brainstorm` | Rapid solution exploration |\n| `/research` | Autonomous research and recommendation |\n| `/audit` | Read-only system mapping |\n| `/retro` | Retrospective analysis |\n| `/decide` | Record key decisions with reasoning |\n| `/hotfix` | Emergency fix escalation |\n\n---\n\n## Architecture\n\n```\nyour-project/\n├── AGENTS.md                    # Global AI governance directives\n├── CLAUDE.md                    # Claude Code integration entry\n│\n├── .agent/                      # Agent Intelligence Layer\n│   ├── config.yaml              # Governance constants\n│   ├── rules/                   # Engineering \u0026 security guardrails\n│   ├── skills/                  # Skill metadata (auto-trigger defs)\n│   └── workflows/               # Workflow definitions (33 workflows)\n│\n├── .agents/skills/              # Full skill implementations (14 skills)\n│   ├── test-driven-development/\n│   ├── systematic-debugging/\n│   └── ...\n│\n├── .agentcortex/                # Runtime \u0026 State Layer\n│   ├── bin/                     # Deploy \u0026 validation scripts\n│   ├── context/                 # SSoT + work logs\n│   ├── docs/                    # Philosophy, platform guides, examples\n│   ├── metadata/                # Skill registry \u0026 cache index\n│   ├── templates/               # Reusable ADR/spec/README templates\n│   └── tools/                   # Runtime tools (Python)\n│\n├── docs/                        # Project-level documentation\n│   ├── specs/                   # Feature specifications\n│   ├── adr/                     # Architecture Decision Records\n│   └── architecture/            # Domain architecture docs\n│\n└── installers/                  # Cross-platform deploy wrappers\n    ├── deploy_brain.sh          # Bash installer\n    ├── deploy_brain.ps1         # PowerShell installer\n    └── deploy_brain.cmd         # CMD installer\n```\n\n---\n\n## Platform Compatibility\n\n| Platform | Status | Integration |\n|:---|:---|:---|\n| **Claude Code** | Native support | `CLAUDE.md` entrypoint + Claude platform guide |\n| **OpenAI Codex** | Native support | `AGENTS.md`, Codex platform guide, and CLI delegation workflow |\n| **Google Antigravity** | Native support | Intent router + Antigravity runtime guidance |\n| **Cursor** | Compatible | Uses `AGENTS.md` / project-rule style guidance |\n| **GitHub Copilot** | Compatible | Uses repository instructions and guardrail docs |\n| **Any LLM Agent** | Compatible | Model-agnostic Markdown workflows and evidence rules |\n\n---\n\n## Philosophy\n\nAgentic OS is built on [10 non-negotiable principles](.agentcortex/docs/AGENT_PHILOSOPHY.md):\n\n| # | Principle | Meaning |\n|:---|:---|:---|\n| 1 | AI Drives, Human Assists | AI follows phases autonomously; human confirms direction |\n| 2 | Never Skip Phases | Sequential execution, not skip-to-end |\n| 3 | Constitution Over Task | Guardrails override any request |\n| 4 | No Evidence = No Completion | Claims without proof are rejected |\n| 5 | Correctness First | Correct \u003e performant \u003e clever |\n| 6 | Token Efficiency | Minimize cold-start cost, maximize context reuse |\n| 7 | Cross-Model Compliance | Works with any AI model, no vendor lock-in |\n| 8 | Actionable Documentation | Every doc must be directly usable |\n| 9 | Scope Discipline | Only touch what was asked |\n| 10 | Explainability | Every AI decision must be traceable |\n\n---\n\n## Documentation\n\nStart with the document that matches what you are trying to do:\n\n| Goal | Start Here | Why |\n|:---|:---|:---|\n| Install or upgrade Agentic OS | [Quick Start](#quick-start), [Migration Guide](.agentcortex/docs/guides/migration.md) | Deployment commands, update flow, and existing-project guidance |\n| Choose the right AI model | [Model Selection Guide](docs/AGENT_MODEL_GUIDE.md), [Lifecycle Benchmark](docs/LIFECYCLE_BENCHMARK.md) | Practical model tradeoffs and measured token costs |\n| Run disciplined agent workflows | [Agent Philosophy](.agentcortex/docs/AGENT_PHILOSOPHY.md), [Testing Protocol](.agentcortex/docs/TESTING_PROTOCOL.md) | Core rules, evidence expectations, and verification standards |\n| Use a specific platform | [Codex Platform Guide](.agentcortex/docs/CODEX_PLATFORM_GUIDE.md), [Claude Platform Guide](.agentcortex/docs/CLAUDE_PLATFORM_GUIDE.md) | Platform-specific loading, handoff, and compatibility notes |\n| Understand governance internals | [Document Governance](.agentcortex/docs/guides/doc-governance.md), [Nonlinear Scenarios](.agentcortex/docs/NONLINEAR_SCENARIOS.md) | State, handoff, recovery, and documentation lifecycle rules |\n| Extend skills safely | [Skill Ecosystem](docs/architecture/skill-ecosystem.md), [Token Governance](.agentcortex/docs/guides/token-governance.md) | Skill packaging, trigger policy, and context budget control |\n\n| Document | Description |\n|:---|:---|\n| [Agent Philosophy](.agentcortex/docs/AGENT_PHILOSOPHY.md) | 10 core principles |\n| [Model Selection Guide](https://github.com/KbWen/agentic-os/blob/main/docs/AGENT_MODEL_GUIDE.md) | Flash vs. Pro guidance |\n| [Testing Protocol](.agentcortex/docs/TESTING_PROTOCOL.md) | Testing standards |\n| [Project Examples](.agentcortex/docs/PROJECT_EXAMPLES.md) | Node.js \u0026 Python examples |\n| [Token Governance](.agentcortex/docs/guides/token-governance.md) | Token optimization strategies |\n| [Token Optimization Quickstart](https://github.com/KbWen/agentic-os/blob/main/docs/guides/token-optimization-quickstart.md) | Reduce token costs immediately ([繁體中文](https://github.com/KbWen/agentic-os/blob/main/docs/guides/token-optimization-quickstart_zh-TW.md)) |\n| [Context Budget](.agentcortex/docs/guides/context-budget.md) | Context window management |\n| [Migration Guide](.agentcortex/docs/guides/migration.md) | Upgrade an existing install |\n| [Codex Platform Guide](.agentcortex/docs/CODEX_PLATFORM_GUIDE.md) | Codex Web/App adaptation |\n| [Claude Platform Guide](.agentcortex/docs/CLAUDE_PLATFORM_GUIDE.md) | Claude Code integration |\n| [Nonlinear Scenarios](.agentcortex/docs/NONLINEAR_SCENARIOS.md) | Recovery from interrupted sessions |\n| [Lifecycle Benchmark](https://github.com/KbWen/agentic-os/blob/main/docs/LIFECYCLE_BENCHMARK.md) | 6 real scenarios with token costs |\n| [生命週期基準測試](https://github.com/KbWen/agentic-os/blob/main/docs/LIFECYCLE_BENCHMARK_zh-TW.md) | Token 消耗量測（繁體中文） |\n\n---\n\n## Contributing\n\nSee [CONTRIBUTING.md](https://github.com/KbWen/agentic-os/blob/main/CONTRIBUTING.md) for guidelines on contributing as a human or AI agent.\n\n## License\n\nMIT License. See [LICENSE](LICENSE) for details.\n\n---\n\n\u003cp align=\"center\"\u003e\n  \u003csub\u003eAn open-source governance layer for AI coding agents. Contributions and feedback are welcome.\u003c/sub\u003e\n\u003c/p\u003e\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkbwen%2Fagentic-os","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fkbwen%2Fagentic-os","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkbwen%2Fagentic-os/lists"}