firstmate: Skill for Claude Code

.agents/skills/firstmate-self-repo/SKILL.md

firstmate-self-repo is a skill for Claude Code from kunchenguid/firstmate. It costs 76 tokens per session (1,130 once invoked), scanned A, original, MIT.

A permission guide for agents working inside the firstmate code repository. It separates normal local use from explicit permission to publish changes to the shared upstream repository.

In plain words
What is it for?
Use it before dispatching, updating, repairing, delivering, or publishing work related to firstmate itself. It also covers granting or removing upstream publication permission.
Why use it?
It prevents an agent from mistaking access to a repository or GitHub account for permission to publish. It helps avoid unexpected branches, pull requests, issues, comments, forks, or reviews.

Skill for Claude Code

Written for Claude Code: user-invocable in frontmatter. Also seen: installed under .agents/ (shared by several agents).

This is kunchenguid/firstmate's own configuration. It tells Claude Code how to work on firstmate itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything firstmate configures →

About the project

firstmate is an agent distribution that lets one coding agent coordinate a crew of autonomous agents working in separate git worktrees. It is for running parallel fixes, investigations, plans, and audits while receiving completed pull requests, merges, or reports through one supervising agent. The catalogue entries provide skills, hooks, instructions, and settings for operating this workflow.

kunchenguid/firstmate · 8,295 stars · on GitHub

Reuse

Borrowing it

Nothing to install: this file belongs to kunchenguid/firstmate. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/kunchenguid/firstmate/main/.agents/skills/firstmate-self-repo/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/kunchenguid/firstmate

Made for: Claude Code.

Wrote 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.

agentmods badge for firstmate-self-repo

README.md
[![agentmods](https://agentmods.dev/badge/skills/kunchenguid/firstmate/firstmate-self-repo/github.svg)](https://agentmods.dev/skills/kunchenguid/firstmate/firstmate-self-repo)
Your own site
<a href="https://agentmods.dev/skills/kunchenguid/firstmate/firstmate-self-repo"><img src="https://agentmods.dev/badge/skills/kunchenguid/firstmate/firstmate-self-repo/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.

agentmods 80×15 button for firstmate-self-repo

Your own site · 80×15
<a href="https://agentmods.dev/skills/kunchenguid/firstmate/firstmate-self-repo"><img src="https://agentmods.dev/badge/skills/kunchenguid/firstmate/firstmate-self-repo.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 76 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,130 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00076 $0.01130
Opus 5.5 $0.00030 $0.00452
Sonnet 5.5 $0.00015 $0.00226
Haiku 4.5 $0.00008 $0.00113

Measured 4d ago against content hash 036c31efd2ad, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

firstmate-self-repo 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 4d 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.

.agents/skills/firstmate-self-repo/SKILL.md · 72 lines

How it starts

The opening of the file, as written. The whole thing — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.

firstmate-self-repo

Every firstmate home clones the same public repository and runs from it, which makes firstmate's own upstream the one repository a running home can always reach. Most homes belong to people who simply use firstmate, not to the people who maintain it, and a running agent that treats "I can see the repo" as "I may publish to it" is how a user's firstmate opens surprise pull requests upstream. This skill owns the boundary between the two relationships.

The default: user/operator

Absent any explicit grant, a home runs as firstmate user/operator. It may install, run, update, reconfigure, and locally repair firstmate, and its ordinary project work flows through the registered delivery modes unchanged. What it may never do on its own is publish to firstmate's upstream: no pushed branch, no pull request, no issue, no comment, no fork, no review.

Role is never inferred. GitHub identity, forge permissions, the name or URL of origin or any other remote, repository ownership, authentication state, environment variables, and CI context all say nothing about role; the only inputs are the explicit local grants below.

When a firstmate bug surfaces in ordinary use

A user/operator home that hits a firstmate bug does local work only:

  1. Diagnose privately - reproduce, isolate, and root-cause inside the home and its task records.
  2. Apply a local workaround when one exists - a config change, a pinned local edit, a different invocation.
  3. Optionally prepare a sanitized local issue draft or a proposed local patch under the task's data directory, for the operator to review and submit by hand.
  4. Report to the captain what happened and what was produced.

A local patch or workaround may touch the gitignored operational surface and a task's own worktree; nothing reaches upstream without the authority below. This posture covers firstmate self-repair exactly the same way it covers a dispatched task: the boundary is about publication, not about which surface holds the fix.

Read the full file on GitHub · 72 lines

Changes

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.

  1. 4d ago Changed · -8 tokens per session 036c31efd2ad
  2. 11d ago First seen · 72 lines · 84 tokens per session scan A b0983223467c

Subscribe to this mod's changes

firstmate-self-repo is a skill published in the GitHub repository kunchenguid/firstmate (8,295 stars, last pushed today), licensed MIT. It adds 76 tokens to every session and 1,130 once invoked, about $0.0003 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 2026-10-14.