agentrig: Skill for Claude Code

.agents/skills/no-self-approve/SKILL.md

no-self-approve is a skill for Claude Code from doidor/agentrig. It costs 30 tokens per session (382 once invoked), scanned A, original, MIT.

A review rule that prevents an agent from approving a pull request (a proposed code change) created by the same shared agent identity.

In plain words
What is it for?
Use it in automated pull-request review and merge checks where approval must come from an independent reviewer.
Why use it?
It avoids invalid self-approval when several agents use one account. The agent can still leave review comments or request changes.

Skill for Claude Code

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

This is doidor/agentrig's own configuration. It tells Claude Code how to work on agentrig 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 agentrig configures →

Reuse

Borrowing it

Nothing to install: this file belongs to doidor/agentrig. 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/doidor/agentrig/main/.agents/skills/no-self-approve/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/doidor/agentrig

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 no-self-approve

README.md
[![agentmods](https://agentmods.dev/badge/skills/doidor/agentrig/no-self-approve.svg)](https://agentmods.dev/skills/doidor/agentrig/no-self-approve)
Your own site
<a href="https://agentmods.dev/skills/doidor/agentrig/no-self-approve"><img src="https://agentmods.dev/badge/skills/doidor/agentrig/no-self-approve.svg" alt="Measured on agentmods" height="20"></a>
Per session 30 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 382 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 $0.00030 $0.00382
Opus 5 $0.00015 $0.00191
Sonnet 5 $0.00006 $0.00076
Haiku 4.5 $0.00003 $0.00038

Measured 6d ago against content hash c3af2f4d7279, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

no-self-approve 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 6d 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/no-self-approve/SKILL.md · 32 lines

What it actually says

no-self-approve (principles 9, 10)

Harness agents often authenticate with one shared bot identity. Casting an approve on a PR that identity authored — including any harness-created PR on an agent/<id>/… branch — is a self-approval: invalid, because approval must come from an independent reviewer (another agent on the same identity cannot supply it either).

Rules

  • Always post your review comments / findings — that work is valuable regardless of who authored the PR.
  • Never cast an approving vote (approve / approve-with-suggestions) on a PR authored by your own identity. If your verdict is "approve" but you authored it, abstain and note that independent approval is required.
  • Negative verdicts (request changes / reject) on your own PR are fine — you may always ask for changes to your own work.

How to check before approving

  • Compare the PR author to your own identity. For GitHub: gh pr view <pr> --json author -q .author.login vs gh api user -q .login.
  • If they match (or the branch is an agent/… branch you produced), do not approve — abstain.
  • Prefer a deterministic pre_merge gate that rejects approver == author, so a slip can never merge. This skill is wired into pre_merge in the state machine.
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. 6d ago First seen · 32 lines · 30 tokens per session scan A c3af2f4d7279

Subscribe to this mod's changes

no-self-approve is a skill published in the GitHub repository doidor/agentrig (5 stars, last pushed 1mo ago), licensed MIT. It adds 30 tokens to every session and 382 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 2026-08-31.