我的代码库

Installation

Switch profiles at install time

bash /tmp/my-codex/install.sh --profile minimal # no packs bash /tmp/my-codex/install.sh --profile dev # data-ai + llmops bash /tmp/my-codex/install.sh --profile full # every installed pack ```

</details>

<details> <summary><strong>Skills — 123 from 4 sources</strong></summary>

Curated per-skill allowlists live in scripts/skill-allowlists.sh — that file is the authority for what ships.

gstack is counted as 26 allowlisted skills plus the repo root entry; the whole gstack repo also lives at ~/.codex/skills/gstack as its canonical runtime tree.

Codex ships no document skills — there is no pdf, docx, pptx, or xlsx skill in this bundle.

</details>

<details> <summary><strong>MCP Servers (3)</strong></summary>

</details>

---

Installation Options

Quick Install

git clone --depth 1 https://github.com/sehoon787/my-codex.git /tmp/my-codex
bash /tmp/my-codex/install.sh
rm -rf /tmp/my-codex

Re-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/.

Full profile (all 2 installed pack categories enabled)

bash /tmp/my-codex/install.sh --profile full ```

config.toml

Global Codex settings in ~/.codex/config.toml:

[agents]
max_threads = 8
max_depth = 1

• max_threads — Maximum concurrent sub-agents
• max_depth — Maximum nesting depth for agent-spawns-agent chains

---

3-Phase Sprint Workflow

For end-to-end feature implementation, Boss orchestrates a structured sprint:

Phase 1: DESIGN         Phase 2: EXECUTE        Phase 3: REVIEW
(interactive)            (autonomous)             (interactive)
─────────────────────   ─────────────────────   ─────────────────────
User decides scope      executor runs tasks     Compare vs design doc
Engineering review      Auto code review        Present comparison table
Confirm "design done"   Architect verification  User: approve / improve

FAQ

<details> <summary><strong>How is my-codex different from my-claude?</strong></summary>

my-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.

</details>

<details> <summary><strong>Can I use both my-codex and my-claude?</strong></summary>

Yes. They install to separate directories (~/.codex/ and ~/.claude/) and do not conflict. Skills from shared upstream sources are adapted for each platform.

</details>

<details> <summary><strong>How do agent packs work?</strong></summary>

Agent packs are domain-specific agent collections installed to ~/.codex/agent-packs/. Two packs ship today — data-ai (13) and llmops (4) — and none is enabled on install. Use my-codex-packs enable <pack> to activate one, or reinstall with --profile full to enable both categories.

</details>

<details> <summary><strong>How does upstream sync work?</strong></summary>

A GitHub Actions workflow runs every 3 days, pulling the latest commits from all 4 upstream submodules, refreshing the SHA pins in upstream/SOURCES.json, and creating a security-gated auto-merge PR. You can also trigger it manually from the Actions tab.

</details>

<details> <summary><strong>What models does my-codex use?</strong></summary>

Boss and sub-orchestrators (Sisyphus, Atlas, Oracle) use gpt-6-astra with high reasoning effort (Boss, Oracle, and Prometheus at xhigh). Standard workers use gpt-5.6-sol with medium reasoning. Lightweight advisory agents use gpt-5.6-terra.

Skills consume the SKILL.md standard as-is with no transformation; only agents are converted to Codex TOML, and the model tier for that conversion is managed from a single file, scripts/model-tiers.sh. When Codex ships its next model generation, update only that file — scripts/md-to-toml.sh and install.sh both source it.

</details>

---

Troubleshooting