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 himraven/codex-chief-engineer --skill chief-engineergit clone --depth 1 https://github.com/himraven/codex-chief-engineerWrote 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/himraven/codex-chief-engineer/chief-engineer)<a href="https://agentmods.dev/skills/himraven/codex-chief-engineer/chief-engineer"><img src="https://agentmods.dev/badge/skills/himraven/codex-chief-engineer/chief-engineer/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/himraven/codex-chief-engineer/chief-engineer"><img src="https://agentmods.dev/badge/skills/himraven/codex-chief-engineer/chief-engineer.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 | $0.00057 | $0.01399 |
| Opus 5.5 | $0.00023 | $0.00560 |
| Sonnet 5.5 | $0.00011 | $0.00280 |
| Haiku 4.5 | $0.00006 | $0.00140 |
Grade A, and why
chief-engineer 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 1mo 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.
How it starts
The opening of the file, as written. The whole thing — 127 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Chief Engineer
The chief owns architecture, contracts, risk, task boundaries, and acceptance. Workers execute bounded decisions. Keep one active chief per objective; provider choice does not transfer authority during a run.
User instructions and existing authorization govern this workflow. Persist necessary decisions and evidence, not conversation history. Reduce repeated context and unnecessary coordination while preserving correctness gates.
Establish the task
- Verify the premise against files, Git state, logs, or current documentation.
- Select the relevant specialist skill and define observable success.
- Keep ambiguity, architecture, state transitions, and red-line decisions with the chief. Delegate bounded evidence gathering and implementation.
- Use the smallest useful topology. A linear task needs no workstreams or phase plan. A single dispatch can reuse the existing task note as chief-state.
- Before choosing an executor, read model routing. Every configurable model starts at high or above; never lower effort to reduce cost. The chief remains responsible for the final decision.
- The chat's selected model runs the chief; executor models are configured separately. This skill cannot switch the current chat's model or retune already-running agents. Verify the actual dispatch configuration, not just the model or Fast setting shown in the main chat.
Approve and delegate
Before write-capable dispatch, obtain explicit human approval for the solution, ownership, executors, execution form, verification, and risks. Wait for it; discussion permits read-only investigation only. A short list is sufficient for one dispatch; use a table when several lanes need comparison.
- Reuse approval already given for this exact scope. Scope, permission, or external-action expansion requires new approval; discussion alone is not it.
- Bind each adapter write brief to a write approval record. An unchanged approved scope does not require asking the user again merely to create that record. Never infer broader authority from a matching hash.
- Keep ship, deploy, publish, and other external actions human-approved.
- Use the standalone brief: outcome, verified context and fixed decisions, owned/forbidden paths, checks, and stop condition. Preserve necessary context; remove raw logs and copied conversation.
- Stop when evidence contradicts the brief, scope expands, or a red line appears. Reslice multiple objectives instead of squeezing them into a brief ceiling.
What ships with it
9 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.
- agents/openai.yaml 224 B
- references/approved-repo-roots.example.txt 287 B
- references/model-routing.md 6.5 KB
- references/operations.md 7.9 KB
- references/review-policy.md 7.3 KB
- references/worker-brief.md 1.6 KB
- references/write-approval.md 726 B
- scripts/ce-dispatch.sh 23 KB runs code
- scripts/ce-token-report.py 27 KB runs code
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.
- 1mo ago Changed · +4 lines 76a9529db505
- 1mo ago Changed · +1 lines 4297cc9e7d59
- 1mo ago Changed · -125 lines · -43 tokens per session a4350daefa68
- 2mo ago First seen · 247 lines · 100 tokens per session scan A 9390ebe56906
chief-engineer is a skill published in the GitHub repository himraven/codex-chief-engineer (5 stars, last pushed 1mo ago), licensed MIT. It adds 57 tokens to every session and 1,399 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-08-31.
Other skills, from other repositories
taskflow
A workflow coordinator for splitting large coding tasks into multiple phases handled by separate agents.
paperclip
Use for Paperclip-managed tasks and heartbeats: reading task context, delivering task documents or files, updating completion or blockers, coordinating or delegating work, and following company governance. Includes control plane API operations for assignments, comments, approvals, and routines.
agentmail
Use your assigned AgentMail inbox to read email tasks, explicitly send or reply, and check delivery. Provided automatically by your inbox assignment.
cw-gates
Use before claiming any Codewhale change is done, green, or ready to land: the focused-to-broad verification ladder, the budget checks CI enforces, and the rules for what counts as a passing test.
feishu
Work with Feishu or Lark bots, docs, sheets, bitables, approval flows, and OpenAPI/MCP setup without hardcoding credentials.
interview
Ask one useful structured question at a time only when material product/implementation choices are genuinely missing; remember answers and produce a brief/spec. Discoverable facts should be investigated instead of asked.