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 jnopareboateng/codex-claude-subagents --skill claude-subagentsgit clone --depth 1 https://github.com/jnopareboateng/codex-claude-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/jnopareboateng/codex-claude-subagents/claude-subagents)<a href="https://agentmods.dev/skills/jnopareboateng/codex-claude-subagents/claude-subagents"><img src="https://agentmods.dev/badge/skills/jnopareboateng/codex-claude-subagents/claude-subagents/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/jnopareboateng/codex-claude-subagents/claude-subagents"><img src="https://agentmods.dev/badge/skills/jnopareboateng/codex-claude-subagents/claude-subagents.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 | $0.00030 | $0.00718 |
| Opus 5 | $0.00015 | $0.00359 |
| Sonnet 5 | $0.00006 | $0.00144 |
| Haiku 4.5 | $0.00003 | $0.00072 |
Grade A, and why
claude-subagents 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 11d 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 — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
claude-subagents
Orchestrate resumable Claude CLI workers as scoped subagents from within Codex.
Overview
Codex is always the lead orchestrator. Claude workers are scoped — each one knows its allowed write path and reports results back via structured logs.
Usage
python3 ~/.codex/skills/claude-subagents/scripts/run_claude_subagent.py \
--task <task-id> \
--prompt <path-to-prompt.md> \
[--write-scope <directory>] \
[--session-id <id>]
| Argument | Required | Description |
|---|---|---|
--task |
yes | Unique task identifier used for log filenames |
--prompt |
yes | Path to the prompt file Codex provides |
--write-scope |
no | Directory the worker is allowed to write to (empty = read-only) |
--session-id |
no | Resume a previous Claude session by ID |
--model |
no | Claude model, default sonnet |
--effort |
no | Reasoning effort, default high |
Logs
All logs are written under .agent-runs/claude/ in the current working directory. This path is automatically added to .gitignore.
| File | Contents |
|---|---|
ledger.json |
Indexed record of all runs (task, session ID, timestamp, exit code) |
<task>.jsonl |
Streaming structured output from claude |
<task>.stderr.log |
stderr from the claude process |
<task>.prompt.md |
The full injected prompt (worker contract + user prompt) |
<task>.summary.md |
Final summary written by the Claude worker |
Worker contract
The launcher injects a preamble into every prompt:
- Codex is lead orchestrator; Claude is a scoped worker.
- Writes are restricted to
--write-scope. - Raw logs are handled by the launcher; the worker must not summarise to stdout.
- The worker must write its final compact summary to
.agent-runs/claude/<task>.summary.md. - The launcher defaults to Sonnet with high reasoning effort.
Concurrency and feedback model
Each launcher call blocks until the Claude worker exits — one worker at a time per call. To run workers in parallel, background multiple launcher calls with distinct --task IDs. Per-task log files do not collide; ledger.json writes are file-locked and use a per-process temp file, so simultaneous completions cannot race or lose an entry. If a session is already in use, the run is marked status: locked and the launcher exits 3.
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.
- 11d ago First seen · 69 lines · 30 tokens per session scan A 9deb4c9b0c62
claude-subagents is a skill published in the GitHub repository jnopareboateng/codex-claude-subagents (15 stars, last pushed 1mo ago), licensed MIT. It adds 30 tokens to every session and 718 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
skill-router
Use when a [skill-router] route card appears in the turn, when no card appeared on a non-trivial task, or when a route looks wrong. Routes every prompt to the right installed skill, pairs it with a process skill, tiers enforcement, and briefs sub-agents. 79 local skills + plugin skills indexed.
psy-exp-reviewer
A review guide for psychological experiment designs, configurations, implementation plans, or code. It checks whether an experiment is logically sound and ready for testing or data collection without changing the materials.
amazing-psycoder
A coordinator for Amazing PsyCoder, a system that turns psychology experiments or experimental data into reviewed code and analysis materials. It sends each request through design, coding, and review stages when the available input requires them.
psy-ana-designer
Design a behavioral-psychology data analysis plan and produce a confirmed analysis config YAML. Use for scientific-question formulation, data-structure intake, statistical-method selection, analysis plans, cleaning/exclusion rules, effect sizes, assumption checks, sensitivity analyses, and figure choices, including…
psy-exp-coder
Generate, modify, or debug runnable psychological experiment code when a completed experiment config YAML or concrete existing code is available. Use for PsychoPy/Python, jsPsych/JavaScript, or Psychtoolbox/MATLAB timing, stimuli, condition files, trial loops, data saving, hardware integration, CJK text, runtime…
psy-ana-coder
Generate, modify, or debug reproducible R or Python analysis code from a completed analysis config YAML or concrete existing script. Use for data import/cleaning, statistical models, assumption checks, effect sizes, sensitivity analyses, publication figures, reports, session information, and analysis-script runtime…