{"id":47640760,"url":"https://github.com/sehoon787/my-codex","last_synced_at":"2026-05-07T11:01:09.245Z","repository":{"id":346306046,"uuid":"1189282140","full_name":"sehoon787/my-codex","owner":"sehoon787","description":"All-in-one agent harness for OpenAI Codex CLI — Boss meta-orchestrator, 400+ agents, 200+ skills, 3 MCP servers. Install once, get every specialist ready.","archived":false,"fork":false,"pushed_at":"2026-05-01T09:32:21.000Z","size":3406,"stargazers_count":11,"open_issues_count":0,"forks_count":2,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-05-01T11:30:41.072Z","etag":null,"topics":["agent-harness","agentic-coding","ai-agents","claude-code","codex","codex-cli","codex-subagents","developer-tools","llm-agent","mcp","mcp-server","multi-agent","openai","orchestration","subagents"],"latest_commit_sha":null,"homepage":"https://sehoon787.github.io/my-codex/","language":"Shell","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/sehoon787.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":"CLA.md"}},"created_at":"2026-03-23T06:50:37.000Z","updated_at":"2026-05-01T09:32:25.000Z","dependencies_parsed_at":"2026-04-02T12:12:33.679Z","dependency_job_id":null,"html_url":"https://github.com/sehoon787/my-codex","commit_stats":null,"previous_names":["sehoon787/my-codex"],"tags_count":117,"template":false,"template_full_name":null,"purl":"pkg:github/sehoon787/my-codex","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sehoon787%2Fmy-codex","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sehoon787%2Fmy-codex/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sehoon787%2Fmy-codex/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sehoon787%2Fmy-codex/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/sehoon787","download_url":"https://codeload.github.com/sehoon787/my-codex/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sehoon787%2Fmy-codex/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":32734391,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-07T02:14:30.463Z","status":"ssl_error","status_checked_at":"2026-05-07T02:14:29.405Z","response_time":62,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.6:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"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-harness","agentic-coding","ai-agents","claude-code","codex","codex-cli","codex-subagents","developer-tools","llm-agent","mcp","mcp-server","multi-agent","openai","orchestration","subagents"],"created_at":"2026-04-02T00:53:03.855Z","updated_at":"2026-05-07T11:01:09.226Z","avatar_url":"https://github.com/sehoon787.png","language":"Shell","funding_links":[],"categories":[],"sub_categories":[],"readme":"[English](./README.md) | [한국어](./docs/i18n/README.ko.md) | [日本語](./docs/i18n/README.ja.md) | [中文](./docs/i18n/README.zh.md) | [Deutsch](./docs/i18n/README.de.md) | [Français](./docs/i18n/README.fr.md)\n\n\u003e [![Claude Code](https://img.shields.io/badge/Claude_Code-my--claude-d97757?style=flat-square\u0026logo=anthropic\u0026logoColor=white)](https://github.com/sehoon787/my-claude) Looking for Claude Code? → **my-claude** — same Boss orchestration in native Claude `.md` agent format\n\n---\n\n\u003cdiv align=\"center\"\u003e\n\n# my-codex\n\n![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)\n![Agents](https://img.shields.io/badge/agents-330%2B-blue)\n![Skills](https://img.shields.io/badge/skills-200%2B-purple)\n![MCP](https://img.shields.io/badge/MCP-3-green)\n![Auto Sync](https://img.shields.io/badge/upstream_sync-weekly-brightgreen)\n\n**All-in-one agent harness for OpenAI Codex CLI.**\n**Install once, 330+ agents ready.**\n\nBoss auto-discovers every agent and skill at runtime,\nthen routes your task to the right specialist via `spawn_agent`. No config. No boilerplate.\n\n\u003cimg src=\"./assets/owl-codex-social.svg\" alt=\"The Maestro Owl — my-codex\" width=\"700\"\u003e\n\n\u003c/div\u003e\n\n---\n\n## Installation\n\n### For Humans\n\n```bash\ncurl -fsSL https://raw.githubusercontent.com/sehoon787/my-codex/main/install.sh | bash\n```\n\nClone-based install:\n\n```bash\ngit clone --depth 1 https://github.com/sehoon787/my-codex.git /tmp/my-codex\nbash /tmp/my-codex/install.sh\nrm -rf /tmp/my-codex\n```\n\nWindows note:\n- `install.sh` patches the npm-managed `codex`, `codex.cmd`, and `codex.ps1` shims when they exist, so the my-codex vault pipeline still has wrapper fallback coverage even if `%APPDATA%\\npm` resolves before `~/.codex/bin`.\n\n### For AI Agents\n\n```\nRead https://raw.githubusercontent.com/sehoon787/my-codex/main/AI-INSTALL.md and follow every step.\n```\n\n---\n\n## How Boss Works\n\nBoss is the meta-orchestrator at the core of my-codex. It never writes code — it discovers, classifies, matches, delegates, and verifies.\n\n```\nUser Request\n     │\n     ▼\n┌─────────────────────────────────────────────┐\n│  Phase 0 · DISCOVERY                        │\n│  Scan ~/.codex/agents/*.toml at runtime     │\n│  → Build live capability registry           │\n└──────────────────────┬──────────────────────┘\n                       ▼\n┌─────────────────────────────────────────────┐\n│  Phase 1 · INTENT GATE                      │\n│  Classify: trivial | build | refactor |     │\n│  mid-sized | architecture | research | ...  │\n│  → Counter-propose skill if better fit      │\n└──────────────────────┬──────────────────────┘\n                       ▼\n┌─────────────────────────────────────────────┐\n│  Phase 2 · CAPABILITY MATCHING              │\n│  P1: Exact skill match                      │\n│  P2: Specialist agent via spawn_agent       │\n│  P3: Multi-agent orchestration              │\n│  P4: General-purpose fallback               │\n└──────────────────────┬──────────────────────┘\n                       ▼\n┌─────────────────────────────────────────────┐\n│  Phase 3 · DELEGATION                       │\n│  spawn_agent with structured instructions   │\n│  TASK / OUTCOME / TOOLS / DO / DON'T / CTX  │\n└──────────────────────┬──────────────────────┘\n                       ▼\n┌─────────────────────────────────────────────┐\n│  Phase 4 · VERIFICATION                     │\n│  Read changed files independently           │\n│  Run tests, lint, build                     │\n│  Cross-reference with original intent       │\n│  → Retry up to 3× on failure               │\n└─────────────────────────────────────────────┘\n```\n\n### Priority Routing\n\nBoss cascades every request through a priority chain until the best match is found:\n\n| Priority | Match Type | When | Example |\n|:--------:|-----------|------|---------|\n| **P1** | Skill match | Task maps to a self-contained skill | `\"merge PDFs\"` → pdf skill |\n| **P2** | Specialist agent | Domain-specific agent exists | `\"security audit\"` → security-reviewer |\n| **P3a** | Boss direct | 2–4 independent agents | `\"fix 3 bugs\"` → parallel spawn |\n| **P3b** | Sub-orchestrator | Complex multi-step workflow | `\"refactor + test\"` → Sisyphus |\n| **P4** | Fallback | No specialist matches | `\"explain this\"` → general agent |\n\n### Model Routing\n\n| Complexity | Model | Used For |\n|-----------|-------|----------|\n| Deep analysis, architecture | o3 (high reasoning) | Boss, Oracle, Sisyphus, Atlas |\n| Standard implementation | o3 (medium) | executor, debugger, security-reviewer |\n| Quick lookup, exploration | o4-mini (low) | explore, simple advisory |\n\n### 3-Phase Sprint Workflow\n\nFor end-to-end feature implementation, Boss orchestrates a structured sprint:\n\n```\nPhase 1: DESIGN         Phase 2: EXECUTE        Phase 3: REVIEW\n(interactive)            (autonomous)             (interactive)\n─────────────────────   ─────────────────────   ─────────────────────\nUser decides scope      executor runs tasks     Compare vs design doc\nEngineering review      Auto code review        Present comparison table\nConfirm \"design done\"   Architect verification  User: approve / improve\n```\n\n---\n\n## Architecture\n\n```\n┌─────────────────────────────────────────────────────┐\n│                    User Request                       │\n└───────────────────────┬─────────────────────────────┘\n                        ▼\n┌─────────────────────────────────────────────────────┐\n│  Boss · Meta-Orchestrator (o3 high)                   │\n│  Discovery → Classification → Matching → Delegation  │\n└──┬──────────┬──────────┬──────────┬─────────────────┘\n   │          │          │          │\n   ▼          ▼          ▼          ▼\n┌──────┐ ┌────────┐ ┌────────┐ ┌────────┐\n│ P3a  │ │  P3b   │ │  P1/P2 │ │Config  │\n│Direct│ │Sub-orch│ │ Skill/ │ │Control │\n│2-4   │ │Sisyphus│ │ Agent  │ │config. │\n│spawn │ │Atlas   │ │ Direct │ │toml    │\n└──────┘ └────────┘ └────────┘ └────────┘\n┌─────────────────────────────────────────────────────┐\n│  Agent Layer (330+ installed TOML files)              │\n│  OMO 9 · OMX 33 · Awesome Core 54 · Superpowers 1   │\n│  + 20 domain agent-packs (on-demand)                  │\n├─────────────────────────────────────────────────────┤\n│  Skills Layer (200+ from ECC + gstack + OMX + more)  │\n│  tdd-workflow · security-review · autopilot           │\n│  pdf · docx · pptx · xlsx · team                     │\n├─────────────────────────────────────────────────────┤\n│  MCP Layer                                            │\n│  Context7 · Exa · grep.app                            │\n└─────────────────────────────────────────────────────┘\n```\n\n---\n\n## What's Inside\n\n| Category | Count | Source |\n|----------|------:|--------|\n| **Core agents** (always loaded) | 98 | Boss 1 + OMO 9 + OMX 33 + Awesome Core 54 + Superpowers 1 |\n| **Agent packs** (on-demand) | 220+ | 20 domain categories from agency-agents + awesome-codex-subagents |\n| **Skills** | 200+ | ECC 180+ · gstack 40 · OMX 36 · Superpowers 14 · Core 2 |\n| **MCP Servers** | 3 | Context7, Exa, grep.app |\n| **config.toml** | 1 | my-codex |\n| **AGENTS.md** | 1 | my-codex |\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eCore Agent — Boss meta-orchestrator (1)\u003c/strong\u003e\u003c/summary\u003e\n\n| Agent | Model | Role | Source |\n|-------|-------|------|--------|\n| Boss | o3 high | Dynamic runtime discovery → capability matching → optimal routing. Never writes code. | my-codex |\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eOMO Agents — Sub-orchestrators and specialists (9)\u003c/strong\u003e\u003c/summary\u003e\n\n| Agent | Model | Role | Source |\n|-------|-------|------|--------|\n| Sisyphus | o3 high | Intent classification → specialist delegation → verification | [oh-my-openagent](https://github.com/code-yeongyu/oh-my-openagent) |\n| Hephaestus | o3 high | Autonomous explore → plan → execute → verify | oh-my-openagent |\n| Atlas | o3 high | Task decomposition + 4-stage QA verification | oh-my-openagent |\n| Oracle | o3 high | Strategic technical consulting (read-only) | oh-my-openagent |\n| Metis | o3 high | Intent analysis, ambiguity detection | oh-my-openagent |\n| Momus | o3 high | Plan feasibility review | oh-my-openagent |\n| Prometheus | o3 high | Interview-based detailed planning | oh-my-openagent |\n| Librarian | o3 medium | Open-source documentation search via MCP | oh-my-openagent |\n| Multimodal-Looker | o3 medium | Image/screenshot/diagram analysis | oh-my-openagent |\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eOMC Agents — Specialist workers (19)\u003c/strong\u003e\u003c/summary\u003e\n\n| Agent | Role | Source |\n|-------|------|--------|\n| analyst | Pre-analysis before planning | [oh-my-claudecode](https://github.com/Yeachan-Heo/oh-my-claudecode) |\n| architect | System design and architecture | oh-my-claudecode |\n| code-reviewer | Focused code review | oh-my-claudecode |\n| code-simplifier | Code simplification and cleanup | oh-my-claudecode |\n| critic | Critical analysis, alternative proposals | oh-my-claudecode |\n| debugger | Focused debugging | oh-my-claudecode |\n| designer | UI/UX design guidance | oh-my-claudecode |\n| document-specialist | Documentation writing | oh-my-claudecode |\n| executor | Task execution | oh-my-claudecode |\n| explore | Codebase exploration | oh-my-claudecode |\n| git-master | Git workflow management | oh-my-claudecode |\n| planner | Rapid planning | oh-my-claudecode |\n| qa-tester | Quality assurance testing | oh-my-claudecode |\n| scientist | Research and experimentation | oh-my-claudecode |\n| security-reviewer | Security review | oh-my-claudecode |\n| test-engineer | Test writing and maintenance | oh-my-claudecode |\n| tracer | Execution tracing and analysis | oh-my-claudecode |\n| verifier | Final verification | oh-my-claudecode |\n| writer | Content and documentation | oh-my-claudecode |\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eAwesome Core Agents (54) — From awesome-codex-subagents\u003c/strong\u003e\u003c/summary\u003e\n\n4 categories installed to `~/.codex/agents/`:\n\n**01-core-development (12)**\naccessibility-tester, ad-security-reviewer, agent-installer, api-designer, code-documenter, code-reviewer, dependency-manager, full-stack-developer, monorepo-specialist, performance-optimizer, refactoring-specialist, tech-debt-analyzer\n\n**03-infrastructure (16)**\nazure-infra-engineer, cloud-architect, container-orchestrator, database-architect, disaster-recovery-planner, edge-computing-specialist, infrastructure-as-code, kubernetes-operator, load-balancer-specialist, message-queue-designer, microservices-architect, monitoring-specialist, network-engineer, serverless-architect, service-mesh-designer, storage-architect\n\n**04-quality-security (16)**\napi-security-tester, chaos-engineer, compliance-auditor, contract-tester, data-privacy-officer, e2e-test-architect, incident-responder, load-tester, mutation-tester, penetration-tester, regression-tester, security-scanner, soc-analyst, static-analyzer, threat-modeler, vulnerability-assessor\n\n**09-meta-orchestration (10)**\nagent-organizer, capability-assessor, conflict-resolver, context-manager, execution-planner, multi-agent-coordinator, priority-manager, resource-allocator, task-decomposer, workflow-orchestrator\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eSuperpowers Agent (1) — From obra/superpowers\u003c/strong\u003e\u003c/summary\u003e\n\n| Agent | Role | Source |\n|-------|------|--------|\n| superpowers-code-reviewer | Comprehensive code review with brainstorming and TDD verification | [superpowers](https://github.com/obra/superpowers) |\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eAgent Packs — On-demand domain specialists (21 categories)\u003c/strong\u003e\u003c/summary\u003e\n\nInstalled to `~/.codex/agent-packs/`. Managed via:\n\n```bash\n# View current state\n~/.codex/bin/my-codex-packs status\n\n# Enable a pack immediately\n~/.codex/bin/my-codex-packs enable marketing\n\n# Switch profiles at install time\nbash /tmp/my-codex/install.sh --profile minimal\nbash /tmp/my-codex/install.sh --profile full\n```\n\n| Pack | Count | Examples |\n|------|------:|---------|\n| engineering | 32 | Backend, Frontend, Mobile, DevOps, AI, Data |\n| marketing | 27 | Douyin, Xiaohongshu, WeChat OA, TikTok, SEO |\n| language-specialists | 27 | Python, Go, Rust, Swift, Kotlin, Java |\n| specialized | 31 | Legal, Finance, Healthcare, Workflow |\n| game-development | 20 | Unity, Unreal, Godot, Roblox, Blender |\n| infrastructure | 19 | Cloud, K8s, Terraform, Docker, SRE |\n| developer-experience | 13 | MCP Builder, LSP, Terminal, Rapid Prototyper |\n| data-ai | 13 | Data Engineer, ML, Database, ClickHouse |\n| specialized-domains | 12 | Supply Chain, Logistics, E-Commerce |\n| design | 11 | Brand, UI, UX, Visual Storytelling |\n| business-product | 11 | Product Manager, Growth, Analytics |\n| testing | 11 | API, Accessibility, Performance, E2E, QA |\n| sales | 8 | Deal strategy, pipeline, outbound |\n| paid-media | 7 | Google Ads, Meta Ads, Programmatic |\n| research-analysis | 7 | Trend, Market, Competitive Analysis |\n| project-management | 6 | Agile, Jira, workflows |\n| spatial-computing | 6 | XR, WebXR, AR/VR, visionOS |\n| support | 6 | Customer support, developer advocacy |\n| academic | 5 | Study abroad, corporate training |\n| product | 5 | Product management, UX research |\n| security | 5 | Penetration testing, compliance, audit |\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eSkills — 200+ from 5 sources\u003c/strong\u003e\u003c/summary\u003e\n\n| Source | Count | Key Skills |\n|--------|------:|------------|\n| [everything-claude-code](https://github.com/affaan-m/everything-claude-code) | 180+ | tdd-workflow, autopilot, security-review, coding-standards |\n| [oh-my-codex](https://github.com/Yeachan-Heo/oh-my-codex) | 36 | plan, team, trace, deep-dive, blueprint, ultrawork |\n| [gstack](https://github.com/garrytan/gstack) | 40 | /qa, /review, /ship, /cso, /investigate, /office-hours |\n| [superpowers](https://github.com/obra/superpowers) | 14 | brainstorming, systematic-debugging, TDD, parallel-agents |\n| [my-codex Core](https://github.com/sehoon787/my-codex) | 2 | boss-advanced, boss-briefing |\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eMCP Servers (3)\u003c/strong\u003e\u003c/summary\u003e\n\n| Server | Purpose | Cost |\n|--------|---------|------|\n| \u003cimg src=\"https://context7.com/favicon.ico\" width=\"16\" height=\"16\" align=\"center\"/\u003e [Context7](https://mcp.context7.com) | Real-time library documentation | Free |\n| \u003cimg src=\"https://exa.ai/images/favicon-32x32.png\" width=\"16\" height=\"16\" align=\"center\"/\u003e [Exa](https://mcp.exa.ai) | Semantic web search | Free 1k req/month |\n| \u003cimg src=\"https://www.google.com/s2/favicons?domain=grep.app\u0026sz=32\" width=\"16\" height=\"16\" align=\"center\"/\u003e [grep.app](https://mcp.grep.app) | GitHub code search | Free |\n\n\u003c/details\u003e\n\n---\n\n## \u003cimg src=\"https://obsidian.md/images/obsidian-logo-gradient.svg\" width=\"24\" height=\"24\" align=\"center\"/\u003e Briefing Vault\n\nObsidian-compatible persistent memory. Every project maintains a `.briefing/` directory that updates during Codex sessions via native plugin hooks, with wrapper fallback for session start/end continuity.\n\n```\n.briefing/\n├── INDEX.md                          ← Project context (auto-created once)\n├── state.json                        ← Session metadata, counters, lastVaultSync (auto-managed)\n├── sessions/\n│   ├── YYYY-MM-DD-\u003ctopic\u003e.md        ← Human/agent-written follow-up session summary\n│   └── YYYY-MM-DD-auto.md           ← Auto-generated scaffold (recorded files, filtered status, follow-up)\n├── decisions/\n│   └── YYYY-MM-DD-\u003cdecision\u003e.md     ← Human/agent-written decision record\n├── learnings/\n│   ├── YYYY-MM-DD-\u003cpattern\u003e.md      ← Human/agent-written learning note\n│   └── YYYY-MM-DD-auto-session.md   ← Auto-generated scaffold (files, wrapper activity, prompts)\n├── references/\n│   └── auto-links.md                ← Reserved for collected research links\n├── agents/\n│   ├── agent-log.jsonl              ← Wrapper/session log\n│   └── YYYY-MM-DD-summary.md        ← Daily logged-signal breakdown\n└── persona/\n    ├── profile.md                   ← Routing/profile summary from logged signals\n    ├── suggestions.jsonl            ← Routing suggestions (auto-generated)\n    ├── persona-policy.json          ← Accepted soft routing preferences for Boss\n    └── rules/                       ← Workflow pattern rules (workflow-*.md)\n```\n\n### Sub-Vaults\n\n| Path | Description |\n|------|-------------|\n| `INDEX.md` | Project overview with links to recent decisions and learnings. Auto-created on first session, refreshed periodically. |\n| `sessions/` | **Session summaries.** `*-auto.md` — auto-generated scaffold refreshed during the session and finalized at stop using recorded session files, filtered status, and logged signals. `\u003ctopic\u003e.md` — human or agent-written follow-up session summary prompted by the vault reminders. |\n| `decisions/` | **Architecture and design decisions** with rationale. Write these as durable notes when a decision is important enough to keep. |\n| `learnings/` | **Patterns, gotchas, non-obvious solutions.** `*-auto-session.md` — auto-generated scaffold refreshed during the session with the session's recorded file list, logged signals, and prompts for follow-up notes. `\u003ctopic\u003e.md` — human or agent-written learning note. |\n| `references/` | **Web research URLs.** `references/auto-links.md` is updated from `WebSearch`/`WebFetch` hook activity when those native Codex hooks are available. |\n| `agents/` | **Logged session signals.** `agent-log.jsonl` — enriched entries with `{ts, agent_id, agent_type, phase, seq, task_hint}`. `YYYY-MM-DD-summary.md` — daily logged-signal breakdown derived from that log. |\n| `persona/` | **User work style profile.** `profile.md` — routing/profile summary derived from logged signals. `suggestions.jsonl` — routing recommendations. `persona-policy.json` — accepted soft routing preferences. `rules/workflow-*.md` — workflow sequence pattern rules proposed by `/boss-briefing`. |\n| `state.json` | Session metadata: counters, lastVaultSync, sessionStartHead. Auto-managed by hooks. |\n\n### Session-Specific Diffs\n\nAt session start, my-codex saves the current git HEAD and a snapshot of the working tree state. During the session, native Codex hooks refresh `.briefing` scaffolds after prompts, edits, searches, and subagent completions. At session end, the final scaffold summarizes diff and status only for recorded paths, while filtering hook-created noise such as `.briefing/` artifacts and session-start `.gitignore` edits.\n\nThis keeps the scaffold focused on session-owned work instead of dumping the entire repository status. For non-git projects, a `YYYY-MM-DD:cwd` identifier is used as fallback.\n\n### Using with Obsidian\n\n1. Open Obsidian → **Open folder as vault** → select `.briefing/`\n2. Notes appear in graph view, linked by `[[wiki-links]]`\n3. YAML frontmatter (`date`, `type`, `tags`) enables structured search\n4. Timeline scaffolds for sessions and learnings build automatically; follow-up summaries, decisions, and learning notes accumulate as you write them\n\n### /boss-briefing\n\nRun `/boss-briefing` during or at the end of a session to:\n- **Sync vault**: Update profile.md, INDEX.md, and agent summaries\n- **Detect workflow patterns**: Analyze temporal agent call sequences across sessions\n- **Recover from gaps**: Generate recovery summaries if days have passed since the last session\n- **Propose persona rules**: Suggest workflow-based routing preferences (not just frequency)\n- **Validate session notes**: Check that today's session has a proper summary\n\nThe Stop hook checks whether `/boss-briefing` has run today. If not, it blocks session end with a reminder. The existing `stop-profile-update.js` continues to run as a fallback.\n\n### Behavioral Hooks\n\n| Hook | Event | Behavior |\n|------|-------|----------|\n| Session Setup | SessionStart | Auto-detects tools + injects Briefing Vault context |\n| Delegation Guard | PreToolUse | Blocks Boss from directly modifying files |\n| Agent Telemetry | PostToolUse | Logs agent usage to analytics |\n| Vault Enforcer | PostToolUse | Counts edits, warns if no vault entries |\n| Subagent Logger | SubagentStop | Logs agent execution to Briefing Vault |\n| Vault Reminder | UserPromptSubmit | Suggests /boss-briefing after 5+ messages |\n| Completion Check | Stop | Runs profile fallback + guards /boss-briefing |\n| Teammate Guide | TeammateIdle | Prompts leader on idle teammates |\n| Quality Gate | TaskCompleted | Verifies deliverable quality |\n\n---\n\n## Upstream Open-Source Sources\n\nmy-codex bundles content from 9 upstream repositories:\n\n| # | Source | What It Provides |\n|---|--------|-----------------|\n| 1 | \u003cimg src=\"https://github.com/sehoon787.png?size=32\" width=\"20\" height=\"20\" align=\"center\"/\u003e **[my-claude](https://github.com/sehoon787/my-claude)** — sehoon787 | Sister project. Same Boss orchestration in native Claude `.md` agent format. Skills, rules, and briefing vault shared across both projects. |\n| 2 | \u003cimg src=\"https://github.com/VoltAgent.png?size=32\" width=\"20\" height=\"20\" align=\"center\"/\u003e **[awesome-codex-subagents](https://github.com/VoltAgent/awesome-codex-subagents)** — VoltAgent | 136 production-grade agents in native TOML format. Already Codex-compatible, no conversion needed. 54 core agents auto-loaded. |\n| 3 | \u003cimg src=\"https://github.com/Yeachan-Heo.png?size=32\" width=\"20\" height=\"20\" align=\"center\"/\u003e **[oh-my-codex](https://github.com/Yeachan-Heo/oh-my-codex)** — Yeachan Heo | 36 skills, hooks, HUD, and team pipelines for Codex CLI. Referenced as architectural inspiration. |\n| 4 | \u003cimg src=\"https://github.com/msitarzewski.png?size=32\" width=\"20\" height=\"20\" align=\"center\"/\u003e **[agency-agents](https://github.com/msitarzewski/agency-agents)** — msitarzewski | 180+ business specialist agent personas across 14 categories. Converted from Markdown to native TOML via automated pipeline. |\n| 5 | \u003cimg src=\"https://github.com/affaan-m.png?size=32\" width=\"20\" height=\"20\" align=\"center\"/\u003e **[everything-claude-code](https://github.com/affaan-m/everything-claude-code)** — affaan-m | 180+ skills across development workflows. Claude Code-specific content stripped; generic coding skills retained. |\n| 6 | \u003cimg src=\"https://github.com/obra.png?size=32\" width=\"20\" height=\"20\" align=\"center\"/\u003e **[superpowers](https://github.com/obra/superpowers)** — Jesse Vincent | 14 skills + 1 agent covering brainstorming, TDD, parallel agents, and code review. |\n| 7 | \u003cimg src=\"https://github.com/code-yeongyu.png?size=32\" width=\"20\" height=\"20\" align=\"center\"/\u003e **[oh-my-openagent](https://github.com/code-yeongyu/oh-my-openagent)** — code-yeongyu | 9 OMO agents (Sisyphus, Atlas, Oracle, etc.). Adapted to Codex-native TOML format. |\n| 8 | \u003cimg src=\"https://github.com/garrytan.png?size=32\" width=\"20\" height=\"20\" align=\"center\"/\u003e **[gstack](https://github.com/garrytan/gstack)** — garrytan | 40 skills for code review, QA, security audit, deployment. Includes Playwright browser daemon. |\n| 9 | \u003cimg src=\"https://github.com/automazeio.png?size=32\" width=\"20\" height=\"20\" align=\"center\"/\u003e **[VibeProxy](https://github.com/automazeio/vibeproxy)** — automazeio | OAuth-based API proxy. Uses existing AI subscriptions (ChatGPT Plus, Gemini, Claude) without separate API keys. Optional setup via `vibeproxy-setup.sh`. |\n\n---\n\n## GitHub Actions\n\n| Workflow | Trigger | Purpose |\n|----------|---------|---------|\n| **CI** | push, PR | Validates TOML agent files, skill existence, and upstream file counts |\n| **Update Upstream** | weekly (Monday) / manual | Runs `git submodule update --remote` and creates auto-merge PR |\n| **Auto Tag** | push to main | Reads version from `config.toml` and creates git tag if new |\n| **Pages** | push to main | Deploys `docs/index.html` to GitHub Pages |\n| **CLA** | PR | Contributor License Agreement check |\n| **Lint Workflows** | push, PR | Validates GitHub Actions workflow YAML syntax |\n\n---\n\n## my-codex Originals\n\nFeatures built specifically for this project, beyond what upstream sources provide:\n\n| Feature | Description |\n|---------|-------------|\n| **Boss Meta-Orchestrator** | Dynamic capability discovery → intent classification → 4-priority routing → delegation → verification |\n| **3-Phase Sprint** | Design (interactive) → Execute (autonomous via executor) → Review (interactive vs design doc) |\n| **Agent Tier Priority** | core \u003e omo \u003e omc \u003e awesome-core deduplication. Most specialized agent wins. |\n| **Cost Optimization** | o4-mini for advisory, o3 for implementation — automatic model routing for 330+ agents |\n| **Briefing Signals** | Wrapper/session logging feeds `.briefing/agents/agent-log.jsonl`, daily summaries, and routing/profile hints |\n| **Smart Packs** | Project-type detection recommends relevant agent packs at session start |\n| **Agent Pack System** | On-demand domain specialist activation via `--profile` and `my-codex-packs` helper |\n| **Codex Attribution** | git hooks record Codex-touched files and append `AI-Contributed-By: Codex` to commit messages |\n| **CI Dedup Detection** | Automated duplicate TOML agent detection across upstream syncs |\n\n---\n\n## Installation Options\n\n### Quick Install\n\n```bash\ngit clone --depth 1 https://github.com/sehoon787/my-codex.git /tmp/my-codex\nbash /tmp/my-codex/install.sh\nrm -rf /tmp/my-codex\n```\n\nRe-running the same command refreshes to the latest `main` build, replaces only my-codex-managed files in `~/.codex/`, and removes stale skill copies from `~/.agents/skills/`.\n\n### Agent Pack Profiles\n\nOn first install, my-codex auto-activates a recommended `dev` set (`engineering`, `design`, `testing`, `marketing`, `support`) and records it in `~/.codex/enabled-agent-packs.txt`.\n\n```bash\n# Minimal profile (core agents only, no packs)\nbash /tmp/my-codex/install.sh --profile minimal\n\n# Full profile (all 21 pack categories enabled)\nbash /tmp/my-codex/install.sh --profile full\n```\n\n### Codex Attribution System\n\n`install.sh` installs a `codex` wrapper plus global git hooks in `~/.codex/git-hooks/`:\n\n- **`prepare-commit-msg`** — Records files changed during a real Codex session\n- **`commit-msg`** — Appends `Generated with Codex CLI: https://github.com/openai/codex` when staged files intersect the recorded change set\n- **`post-commit`** — Adds `AI-Contributed-By: Codex` trailer to qualifying commits\n\nOpt-in `Co-authored-by` trailer: set both `git config --global my-codex.codexContributorName '\u003clabel\u003e'` and `my-codex.codexContributorEmail '\u003cgithub-linked-email\u003e'`. Disable entirely: `git config --global my-codex.codexAttribution false`. my-codex does **not** change `git user.name`, `git user.email`, or commit author identity.\n\n### Agent TOML Format\n\nEvery agent is a native TOML file in `~/.codex/agents/`:\n\n```toml\nname = \"debugger\"\ndescription = \"Focused debugging specialist — traces failures to root cause\"\nmodel = \"o3\"\nmodel_reasoning_effort = \"medium\"\n\n[developer_instructions]\ncontent = \"\"\"\nYou are a debugging specialist. Analyze failures systematically:\n1. Reproduce the issue\n2. Isolate the root cause\n3. Propose a minimal fix\n4. Verify the fix does not break adjacent behavior\n\"\"\"\n```\n\n### config.toml\n\nGlobal Codex settings in `~/.codex/config.toml`:\n\n```toml\n[agents]\nmax_threads = 8\nmax_depth = 1\n```\n\n- `max_threads` — Maximum concurrent sub-agents\n- `max_depth` — Maximum nesting depth for agent-spawns-agent chains\n\n---\n\n## Bundled Upstream Versions\n\nUpstream sources managed as git submodules. Pinned commits tracked in `.gitmodules`.\n\n| Source | Sync |\n|--------|------|\n| [agency-agents](https://github.com/msitarzewski/agency-agents) | submodule |\n| [everything-claude-code](https://github.com/affaan-m/everything-claude-code) | submodule |\n| [oh-my-codex](https://github.com/Yeachan-Heo/oh-my-codex) | submodule |\n| [awesome-codex-subagents](https://github.com/VoltAgent/awesome-codex-subagents) | submodule |\n| [gstack](https://github.com/garrytan/gstack) | submodule |\n| [superpowers](https://github.com/obra/superpowers) | submodule |\n\n---\n\n## FAQ\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eHow is my-codex different from my-claude?\u003c/strong\u003e\u003c/summary\u003e\n\nmy-codex and my-claude share the same Boss orchestration architecture and upstream skill sources. The key difference is the runtime: my-codex targets OpenAI Codex CLI with native `.toml` agent format and `spawn_agent` delegation, while my-claude targets Claude Code with `.md` agent format and the Agent tool.\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eCan I use both my-codex and my-claude?\u003c/strong\u003e\u003c/summary\u003e\n\nYes. They install to separate directories (`~/.codex/` and `~/.claude/`) and do not conflict. Skills from shared upstream sources are adapted for each platform.\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eHow do agent packs work?\u003c/strong\u003e\u003c/summary\u003e\n\nAgent packs are domain-specific agent collections installed to `~/.codex/agent-packs/`. On first install, a `dev` profile is auto-activated. Use `my-codex-packs enable \u003cpack\u003e` to activate additional packs, or reinstall with `--profile full` to enable all 21 categories.\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eHow does upstream sync work?\u003c/strong\u003e\u003c/summary\u003e\n\nA GitHub Actions workflow runs every Monday, pulling the latest commits from all upstream submodules and creating an auto-merge PR. You can also trigger it manually from the Actions tab.\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong\u003eWhat models does my-codex use?\u003c/strong\u003e\u003c/summary\u003e\n\nBoss and sub-orchestrators (Sisyphus, Atlas, Oracle) use o3 with high reasoning effort. Standard workers use o3 with medium reasoning. Lightweight advisory agents use o4-mini.\n\n\u003c/details\u003e\n\n---\n\n## Troubleshooting\n\n### Skills-only recovery\n\nIf a tool reports invalid `SKILL.md` files under `~/.agents/skills/`, the most common cause is a stale local copy or stale symlink target from an older install.\n\nRemove the affected directories from `~/.agents/skills/` and matching entries under `~/.claude/skills/`, then reinstall:\n\n```bash\nnpx skills add sehoon787/my-codex -y -g\n```\n\nIf you use the full Codex bundle, rerun `install.sh` once as well. The full installer refreshes `~/.codex/skills/` and removes stale my-codex-managed copies under `~/.agents/skills/`.\n\n---\n\n## Contributing\n\nIssues and PRs are welcome. When adding a new agent, add a `.toml` file to `codex-agents/core/` or `codex-agents/omo/` and update the agent list in `SETUP.md`. See [CONTRIBUTING.md](./CONTRIBUTING.md) for PR validation steps and Codex commit attribution behavior.\n\n## Credits\n\nBuilt on the work of: [my-claude](https://github.com/sehoon787/my-claude) (sehoon787), [awesome-codex-subagents](https://github.com/VoltAgent/awesome-codex-subagents) (VoltAgent), [oh-my-codex](https://github.com/Yeachan-Heo/oh-my-codex) (Yeachan Heo), [agency-agents](https://github.com/msitarzewski/agency-agents) (msitarzewski), [everything-claude-code](https://github.com/affaan-m/everything-claude-code) (affaan-m), [oh-my-openagent](https://github.com/code-yeongyu/oh-my-openagent) (code-yeongyu), [gstack](https://github.com/garrytan/gstack) (garrytan), [superpowers](https://github.com/obra/superpowers) (Jesse Vincent), [openai/skills](https://github.com/openai/skills) (OpenAI).\n\n## License\n\nMIT License. See the [LICENSE](./LICENSE) file for details.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsehoon787%2Fmy-codex","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsehoon787%2Fmy-codex","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsehoon787%2Fmy-codex/lists"}