Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add WildPigKing/codex-project-skill-manager --skill manage-codex-skillsgit clone --depth 1 https://github.com/WildPigKing/codex-project-skill-managerWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/wildpigking/codex-project-skill-manager/manage-codex-skills)<a href="https://agentmods.dev/skills/wildpigking/codex-project-skill-manager/manage-codex-skills"><img src="https://agentmods.dev/badge/skills/wildpigking/codex-project-skill-manager/manage-codex-skills/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/wildpigking/codex-project-skill-manager/manage-codex-skills"><img src="https://agentmods.dev/badge/skills/wildpigking/codex-project-skill-manager/manage-codex-skills.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00058 | $0.00370 |
| Opus 5 | $0.00029 | $0.00185 |
| Sonnet 5 | $0.00012 | $0.00074 |
| Haiku 4.5 | $0.00006 | $0.00037 |
Grade A, and why
manage-codex-skills scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Manage Codex Skills
Treat skills-menu.json as the registry source and SKILLS_MENU.md as generated output.
Route a task
- Scan every menu row by skill name, trigger, and output.
- For each distinct task or subrequest, shortlist at most three relevant skills.
- Read only the complete SKILL.md files for the shortlist.
- State the order when multiple skills are complementary.
- Continue without a skill when there is no good match.
- Do not bulk-load all skills or route by category name alone.
Organize a library
- Inventory skill names, descriptions, paths, categories, sources, and statuses read-only.
- Detect missing registrations, duplicate names, invalid frontmatter, and path conflicts.
- Propose a category for each unclassified skill with a short reason and confidence. This proposal is read-only and is not a persisted classification.
- Show exact directory creations and optional relocations.
- Let the user approve all, some, or none. An unmanaged skill stays in place and remains usable.
- Apply only approved changes. Recording a category, rebuilding a user menu, or moving a skill is a write. Then validate the resulting menu.
Run scripts/rebuild_skills_menu.py to generate the menu. Run scripts/validate_skills_menu.py before reporting completion.
Passing format validation does not prove a skill is effective. Use draft before representative testing, tested after one successful realistic use, and stable only after repeated successful use.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 31 lines · 58 tokens per session scan A bdbbae2e3312
manage-codex-skills is a skill published in the GitHub repository WildPigKing/codex-project-skill-manager (22 stars, last pushed 14d ago), licensed MIT. It adds 58 tokens to every session and 370 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
codex-prompt
Generate structured, self-contained Codex task prompts. Use when the user asks for a Codex prompt, delegation task, fix, build, audit, ops/tuning, refactor, PR-series, or master-session brief to hand to another Codex run. Grounds every cited path and command, distinguishes directive from investigative work, and binds…
retro-distill
Run a self-distillation retrospective on recent OpenAI Codex sessions (or any Agent-Skills-compatible agent). Use when the user says 'retro,' 'retrospective,' 'distill,' 'what did we learn,' 'review sessions,' 'self-improve,' or after completing a major chunk of work. Covers three modes: quick retro (single session)…
gate-engineering
Design, implement, review, or repair durable quality gates such as linters, CI checks, tripwires, corpus detectors, contract suites, and snapshot or golden tests. Use when asked to add a check that catches a defect, when a recurring defect needs class-wide prevention, when a gate is being trusted or recalibrated, or…
git-under-contention
Operate Git safely when multiple humans or coding agents may share a repository, checkout, index, branches, or worktrees. Use before staging, committing, pushing, merging, rebasing, creating or removing worktrees, recovering apparently lost edits, handling staged files you did not touch, or preserving unfinished agent…
session-close
Close a coding-agent session with durable repository state, proportional verification, documentation reconciliation, and a useful handoff. Use when the user says close out, wrap up, end session, session report, doc sync, preserve work, or asks for a final commit and push before stopping.
verify-honestly
Choose proportional verification and write evidence-bounded completion claims for coding-agent work. Use before saying work is fixed, done, working, passing, shipped, or deployed; when asked to verify a change; when relying on another agent's findings; or when reporting counts, artifact properties, UI behavior, live…