moonbridge: Skill for Claude Code

.agents/skills/agent-bot-identity/SKILL.md

agent-bot-identity is a skill for Claude Code, Codex from briandconnelly/moonbridge. It costs 113 tokens per session (7,764 once invoked), scanned A, a copy of agent-bot-identity, MIT.

Guidelines for giving a local coding agent its own GitHub App identity. A GitHub App identity makes the agent’s commits, pushes, and pull requests appear under a bot account while leaving the developer’s personal Git setup unchanged.

In plain words
What is it for?
Configuring or reviewing bot identities, per-project routing, organization-based routing, and human-authored commits made through an agent.
Why use it?
It separates human and agent attribution, making it clear who performed each repository action without changing the user’s normal credentials.

Skill for Claude CodeCodex

Written for Claude Code and Codex: SessionStart hook event, but also agents/openai.yaml present. Also seen: reads .claude/ paths; mentions subagents; mentions Claude Code.

This is briandconnelly/moonbridge's own configuration. It tells Claude Code and Codex how to work on moonbridge 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 moonbridge configures →

Part of the moonbridge plugin — 5 skills, 1 MCP server shipped together

Reuse

Borrowing it

Nothing to install: this file belongs to briandconnelly/moonbridge. 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/briandconnelly/moonbridge/main/.agents/skills/agent-bot-identity/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/briandconnelly/moonbridge

Made for: Claude Code, Codex.

Or install moonbridge, the plugin that ships this one along with the rest of its 5 skills, 1 MCP server.

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 agent-bot-identity

README.md
[![agentmods](https://agentmods.dev/badge/skills/briandconnelly/moonbridge/agent-bot-identity/github.svg)](https://agentmods.dev/skills/briandconnelly/moonbridge/agent-bot-identity)
Your own site
<a href="https://agentmods.dev/skills/briandconnelly/moonbridge/agent-bot-identity"><img src="https://agentmods.dev/badge/skills/briandconnelly/moonbridge/agent-bot-identity/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 agent-bot-identity

Your own site · 80×15
<a href="https://agentmods.dev/skills/briandconnelly/moonbridge/agent-bot-identity"><img src="https://agentmods.dev/badge/skills/briandconnelly/moonbridge/agent-bot-identity.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 113 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 7,764 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.
Origin 100% copy Near-identical to another mod 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 $0.00113 $0.07764
Opus 5 $0.00056 $0.03882
Sonnet 5 $0.00023 $0.01553
Haiku 4.5 $0.00011 $0.00776

Measured 9d ago against content hash 2af69564acb8, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

agent-bot-identity scanned grade A with 1 finding 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 9d ago.

The scan reads SKILL.md. This mod also ships 4 executable files (scripts/claude/bot-env-hook.sh, scripts/claude/session-env.sh, tests/selector-cache-test.sh, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

Asks for rootlowPrivilege escalation

A mod that escalates privileges can change anything on the machine, not only the project.

6. Generate a private key; store at `~/.config/acme-agent/key.pem`, `chmod 600`.

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Origin

This is a copy

100% identical to agent-bot-identity — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

.agents/skills/agent-bot-identity/SKILL.md · 299 lines

How it starts

The opening of the file, as written. The whole thing — 299 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Bot Identity

Core is harness-neutral; adapters ship for Claude Code (tested) and Codex CLI — Variant A partially verified with its GitHub write path not exercised and a documented as-me limitation, Variant B pending — see Phase 4.

Overview

Give a local coding agent its own GitHub App identity so the commits, pushes, and PRs it makes attribute to a bot by default, while the human's git setup (SSH key, GPG signing, keychain credentials) stays untouched on the same machine. Attribution is per unit of work: autonomous agent work carries the bot identity, and in repos where the human also contributes through the agent, collaborated commits can carry the human's authorship via the as-me wrapper (Phase 3) while auth still rides the bot token. The isolation mechanism is local routing that injects the bot's credentials only into the agent's sessions where the bot belongs — either per-project opt-in or automatic org-gated routing, depending on the harness adapter (Phase 4) — without editing any shell dotfiles.

Core principle: this buys attribution, not containment. The per-project scoping makes the well-behaved default path use the bot identity; the only hard boundaries are the App's installation list — which bounds git/content access and all access to private repos, not everything (see What This Enforces) — and the server-side rulesets of the repos it touches. Never present this setup as a sandbox.

When to Use

  • Setting up Claude Code (or a similar local agent) to commit, push, and open PRs as a bot on a developer machine, with personal git operations unchanged.
  • Splitting attribution in a repo where you contribute both directly and via the agent — collaborated work authored as you, autonomous work as the bot, without separate checkouts or settings toggles.
  • Auditing an existing dual-identity setup for over-trust — e.g. a review gate assumed to bind the agent, or local scoping treated as a security boundary.
  • Symptoms: agent commits show the human as author; "have the agent open PRs as a bot"; bot PRs need human approval but the agent runs on the human's laptop.

Read the full file on GitHub · 299 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. 9d ago First seen · 299 lines · 113 tokens per session scan A 2af69564acb8

Subscribe to this mod's changes

agent-bot-identity is a skill published in the GitHub repository briandconnelly/moonbridge (0 stars, last pushed today), licensed MIT. It adds 113 tokens to every session and 7,764 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 1 finding (asks for root). It is 100% identical to agent-bot-identity, differing in 0 lines, and is treated as a copy.