Borrowing it
Nothing to install: this file belongs to leonardsellem/codex-specialized-subagents. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/leonardsellem/codex-specialized-subagents/master/.codex/skills/delegation-autopilot/SKILL.mdgit clone --depth 1 https://github.com/leonardsellem/codex-specialized-subagentsWrote 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/leonardsellem/codex-specialized-subagents/delegation-autopilot)<a href="https://agentmods.dev/skills/leonardsellem/codex-specialized-subagents/delegation-autopilot"><img src="https://agentmods.dev/badge/skills/leonardsellem/codex-specialized-subagents/delegation-autopilot/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/leonardsellem/codex-specialized-subagents/delegation-autopilot"><img src="https://agentmods.dev/badge/skills/leonardsellem/codex-specialized-subagents/delegation-autopilot.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 | $0.00035 | $0.00700 |
| Opus 5 | $0.00017 | $0.00350 |
| Sonnet 5 | $0.00007 | $0.00140 |
| Haiku 4.5 | $0.00003 | $0.00070 |
Grade A, and why
delegation-autopilot 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.
How it starts
The opening of the file, as written. The whole thing — 56 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Delegation autopilot (parent-agent only)
This skill is meant for the parent Codex agent in interactive mode.
Prerequisites (avoid common first-run failures)
Delegated runs can take minutes. If tool calls fail around ~60s with an error like timed out awaiting tools/call after 60s, increase the MCP tool timeout for this server in ${CODEX_HOME:-~/.codex}/config.toml:
[mcp_servers.codex-specialized-subagents]
tool_timeout_sec = 600
When to call delegate_autopilot
- The user request is multi-step ("and", "then", "also", "plus") or touches multiple areas (code + tests + docs).
- The change is cross-cutting (multiple files/modules) or likely needs specialist attention (security/perf/research).
- The user explicitly asks to delegate / use subagents.
Default rules:
- If the request includes code changes and (tests or docs), call
delegate_autopilotbefore running shell commands or editing files. - If the request is multi-step research-only and you want consistent “scan/implement/verify” decomposition or per-job
thinking_levelbehavior, calldelegate_autopilotwithsandbox="read-only".
When not to call
- The user is asking a simple question or wants a short explanation.
- The change is tiny and local (single file, trivial edit) and you can do it directly.
How to call (minimal)
Call MCP tool delegate_autopilot with:
task: the user request (verbatim)cwd: current workspace directory (optional; defaults are OK)
Optional knobs:
sandbox: use"workspace-write"when you want subagents to edit files;"read-only"for analysis-only delegation; avoid"danger-full-access"unless absolutely necessary.max_agents: default3(use1when you want a single subagent run).skills_mode: default"auto"(use"none"to disable skill selection).
After calling
- If
structuredContent.decision.should_delegateisfalse: continue normally without delegation. - Otherwise: use
structuredContent.aggregateas the consolidated result, and inspectstructuredContent.run_dirfor artifacts.
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 · 56 lines · 35 tokens per session scan A 241055a0c1d7
delegation-autopilot is a skill published in the GitHub repository leonardsellem/codex-specialized-subagents (68 stars, last pushed 8mo ago), licensed MIT. It adds 35 tokens to every session and 700 once invoked, about $0.0002 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
html-portal-generator
Convert a codebase into a self-contained HTML portal app for ingestion into AI application systems. Produces a single deployable HTML file with embedded CSS, JS, and data.
api-designer
Design RESTful APIs with best practices for consistency and usability.
coding-assistant
Provides coding assistance with best practices and code review.
debug-helper
Systematic debugging approach for identifying and fixing issues.
test-writer
Write thorough tests following TDD and BDD principles.
git-workflow
Git best practices for version control and collaboration.