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 metraton/gaia --skill subagent-request-approvalgit clone --depth 1 https://github.com/metraton/gaiaWrote 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/metraton/gaia/subagent-request-approval)<a href="https://agentmods.dev/skills/metraton/gaia/subagent-request-approval"><img src="https://agentmods.dev/badge/skills/metraton/gaia/subagent-request-approval.svg" alt="Measured on agentmods" 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 | $0.00026 | $0.00520 |
| Opus 5 | $0.00013 | $0.00260 |
| Sonnet 5 | $0.00005 | $0.00104 |
| Haiku 4.5 | $0.00003 | $0.00052 |
Grade A, and why
subagent-request-approval 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 7d 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 — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Request Approval — Producer Branch
Plan-first set
After read-only investigation, collect every predictable exact T3 command that forms one coherent bounded operation -- one command or several. Do not attempt the commands first, even a single one: request it proactively through the same verb.
Run one CLI request, with one --command per atomic item in execution order:
gaia approvals request-set --command '<exact 0>' --command '<exact 1>' --rationale '<goal, risk, rollback, verification>' --agent-id <agent_id> --session-id <session_id>
A single predictable T3 command uses the identical verb with one --command.
The CLI validates T3 eligibility, persists REQUESTED, and returns the
approval_id, ordered set, and fingerprints. Relay those returned values inside
approval_request; do not calculate an id or fingerprint yourself.
Blocked single command
When PreToolUse returns [T3_BLOCKED], stop. Copy the returned approval_id
and sealed payload verbatim into approval_request. Do not retry, reword, split,
wrap, or seek the same effect through another tool.
Checkpoint and stop
- Set
agent_statetoAPPROVAL_REQUEST. - Put all exact consent data in
approval_request, including full ordered set, risk, rollback, and verification when COMMAND_SET. - Add only the block/request outcome to evidence; do not duplicate payloads.
- Set
pending_stepsto execution after consent andnext_actionto relay the request to the user. - Finalize the non-terminal contract, emit its compatibility fence, and stop.
The producer does not present an approval as already granted and does not verify
execution before it happens. orchestrator-present-approval presents consent;
after grant, a fresh specialist dispatch loads execution.
Grouping boundary
Use a set only for exact commands already implied by the accepted plan, with one risk/rollback/verification narrative. Do not group unrelated effects, alternatives, speculative cleanup, or commands derived from earlier outputs. Consent grouping is not atomic execution.
What ships with it
1 file 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.
- 7d ago First seen · 57 lines · 26 tokens per session scan A 019c1a4dd988
subagent-request-approval is a skill published in the GitHub repository metraton/gaia (3 stars, last pushed today), licensed MIT. It adds 26 tokens to every session and 520 once invoked, about $0.0001 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 2026-08-31.
Other skills, from other repositories
review-team
A multi-reviewer code review process that checks a change from several specialist viewpoints and combines the results into one report. It can cover bugs, security, tests, dependencies, frontend behavior, and continuous-integration workflows.
ultra
Fans the work out as a fleet of parallel Grok and Codex agents billed to their own subscriptions, then synthesizes one result. The peer engine equivalent of ultracode, adding intensity without spending Claude quota on the fleet. Use it for genuinely broad goals, not only explicit asks for intensity.
grok-prompting
Brief writing guidance for composing self contained Grok briefs for coding, review, diagnosis, and second opinion tasks.
smoke
Runs a three probe live smoke wave after a plugin update and reports gate chain health before real work rides it.
codex-result-handling
Internal contract for returning Codex companion output without alteration.
shipwrights-loop
Drive multiple Jira tickets sequentially through the /shipwrights-epic pipeline. Auto-picks the next ticket, watches the PR until merged, transitions Jira to shipped, repeats. Resumable across Claude Code sessions via .shipwrights/loop-state.json. Invoked as /shipwrights-loop [N] | --status | --abort.