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 escoffier-labs/skillet --skill stationsgit clone --depth 1 https://github.com/escoffier-labs/skilletWrote 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/escoffier-labs/skillet/stations)<a href="https://agentmods.dev/skills/escoffier-labs/skillet/stations"><img src="https://agentmods.dev/badge/skills/escoffier-labs/skillet/stations/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/escoffier-labs/skillet/stations"><img src="https://agentmods.dev/badge/skills/escoffier-labs/skillet/stations.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 | $0.00062 | $0.01054 |
| Opus 5 | $0.00031 | $0.00527 |
| Sonnet 5 | $0.00012 | $0.00211 |
| Haiku 4.5 | $0.00006 | $0.00105 |
Grade A, and why
stations 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 — 50 lines — stays where its author put it; the contents beside it link to each section on GitHub.
stations
A kitchen runs fast because each station cooks its own dish: garde manger does not wait for the grill. But the speed comes from the cuts being made before service, and two cooks at one board ruin both dishes. This skill is the expeditor's fan-out: deciding what is genuinely separate work, giving each station a complete ticket, and tasting everything together before it leaves.
Core principle: parallelism is earned by proving independence, not assumed from the symptom list. The fan-out is the last step of triage, never the first move.
Triage before any dispatch
The expensive mistake is dispatching one agent per symptom. Symptoms group by root cause, and a refactor that broke four things usually broke two things twice.
- Cluster by suspected cause, not by file. Two failures in different test files that both touch the same layer (the same data shape, the same recent change) are one cluster until proven otherwise. Check what actually changed (
git log, the refactor's diff) before declaring anything independent. Symptoms living in different files is not independence; it is the same bug crashing in two places. - One cluster, one station. A shared root cause dispatched to two agents buys two local symptom patches that each make their own test green and disagree with each other - one patches the data producer, the other patches a consumer, and the suite is green while the design is incoherent. The cluster goes to a single agent with the whole picture, applying refire's root-cause discipline inside.
- Check the write sets. Stations that would edit the same files do not run concurrently: serialize them, or give each its own worktree (prefer the harness's native worktree tool; verify a manual worktree directory is gitignored). Read-only overlap is fine.
- Anything exploratory stays with you. If you do not yet know what is broken, that investigation is not parallelizable; fan out after it, not instead of it.
When in doubt, fewer stations. A wrongly merged cluster costs one slightly bigger ticket; a wrongly split one ships two half-fixes.
What ships with it
2 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 · 50 lines · 62 tokens per session scan A 370069f71394
stations is a skill published in the GitHub repository escoffier-labs/skillet (4 stars, last pushed 8d ago), licensed MIT. It adds 62 tokens to every session and 1,054 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 2026-08-31.
Other skills, from other repositories
assess-interview-candidate
A structured hiring-assessment workflow that turns a candidate's résumé and a job description into reviewable evidence, interview questions, and an offline HTML report for interviewers.
task-clarifier
Deep need-clarification skill. Use only when the user explicitly invokes $task-clarifier. Once activated, keep asking until all three goals are met: the user fully understands their own needs, the AI fully understands the user's needs, and the user confirms the AI's understanding is correct. Do not intervene in task…
user-profile-keeper
Local user-profile maintenance skill for Codex, Claude Code, OpenClaw, OpenCode, and other agent harnesses. Use only when the user explicitly invokes $user-profile-keeper to create, initialize, update, query, correct, delete, export, or audit a local persistent user profile. It can extract durable collaboration…
run-history-skill-builder
Turn a completed task, browser flow, artifact pipeline, failure-recovery trace, or repeatedly refined workflow into a new reusable skill package or a reviewed skill-design plan. Use when the user asks to make a new skill from real run history, extract a reusable workflow from conversation/logs/files, summarize lessons…
run-history-skill-upgrader
Use real run evidence, validation failures, source drift, platform drift, and user feedback to plan and, only after explicit approval, apply structural upgrades to an existing skill. Use when the user asks to improve an existing skill from recent runs, recurring failures, outdated sources, excessive bloat, changed…
session-handoff-prompt
Create a concise continuation prompt that a fresh agent session can paste in to resume a long or degraded session. Use when the user asks for a handoff prompt, restart prompt, continuation prompt, context transfer, fresh-session resume, or a compact summary for opening a new session. Do not use for ordinary summaries…