codex-maintainer-kit CLAUDE.md

A set of instructions for maintaining a repository with a staged, multi-agent pull-request workflow. It identifies the main files, checks, and rules that agents should follow.

In plain words
What is it for?
It is for navigating the repository, running its quality and contract checks, coordinating review stages, and creating audit-ready changes.
Why use it?
It gives coding agents a clear starting point and shared process for reviewing, validating, and preparing changes.

Instructions file

Install

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.

agentmods
npx agentmods add instructions/atman36/codex-maintainer-kit/claude-md
Clone the repo
git clone --depth 1 https://github.com/Atman36/codex-maintainer-kit
Per session 939 This file is loaded in full into every session.
When invoked 939 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
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 $0.00939 $0.00939
Opus 5 $0.00469 $0.00469
Sonnet 5 $0.00188 $0.00188
Haiku 4.5 $0.00094 $0.00094

Measured 2d ago against content hash 3ec8ebffea4b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

codex-maintainer-kit CLAUDE.md 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 2d 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.

CLAUDE.md · 66 lines

How it starts

The opening of the file, as written. The whole thing — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Codex Maintainer Kit — Project Instructions

Overview

Multi-agent PR pipeline: Scout → Analyst → Critic → Gatekeeper → Implementer → Reviewer → PR Writer → Publisher.

Fastest Entrypoint

  • Start with program.md when you want the shortest agent-facing file for this repository.
  • Use it to understand what to read first, which workflow is primary, and which constraints are mandatory.

Skills

See AGENTS.md for the full list of available skills and trigger rules.

Key Files

  • program.md — shortest agent-facing entrypoint for the repo
  • tools/quality_gate.py — forbidden-files scanner + merge-probability heuristic + autoclean (CLI: --autoclean-unplanned)
  • tools/run_pipeline.py — deterministic DAG orchestrator (stage order + gate checks + retries)
  • tools/contract_registry.py — canonical placeholder registry + static contract validation
  • tools/validate_skills.py — CI contract check for skills
  • tools/make_agent_audit_archive.py — builds clean audit archives (excludes nested junk)
  • schemas/execution_result.schema.jsonExecutionResult envelope (all stage outputs must conform)
  • schemas/prspec.schema.jsonPRSpec schema (validated at runtime in pipeline)
  • tools/tests/ — regression tests (run with python3 -m unittest discover -s tools/tests -p 'test_*.py')
  • config/pipeline_modes.json — pipeline mode analysis graphs (loaded at startup, early-validated)
  • pyproject.toml — minimal editable-install packaging metadata (pip install -e .)
  • analysis_report/ — stable repo-owned artifact directory for pipeline summaries and per-run outputs

Placeholders

Canonical set (enforced by contract_registry.py) — key ones:

  • {{REPO_ROOT}}, {{REPO_URL}}, {{BASE_BRANCH}}, {{HEAD_BRANCH}}
  • {{CONSTRAINTS}}, {{ALLOWED_COMMANDS}}
  • {{ARTIFACT_DIR}} — where analysis-stage JSON artifacts are written (default: analysis_report/)
  • {{REPORT_PATH}} — where a markdown report should be written (default: analysis_report/runs/<run_id>/report.md)
  • {{MODE}}, {{FOCUS}}, {{STAGE}}, {{COMMAND}}, {{RUNNER}}
  • Per-stage JSON paths: {{SCOUT_JSON}}, {{ANALYST_JSON}}, {{ARCHITECT_JSON}}, {{CRITIC_JSON}}, {{GATEKEEPER_JSON}}, {{IMPLEMENT_JSON}}, {{REVIEWER_JSON}}, {{PRSPEC_JSON}}

Read the full file on GitHub · 66 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. 2d ago First seen · 66 lines · 939 tokens per session scan A 3ec8ebffea4b

Subscribe to this mod's changes

codex-maintainer-kit CLAUDE.md is an instructions file published in the GitHub repository Atman36/codex-maintainer-kit (2 stars, last pushed 3mo ago), licensed MIT. It adds 939 tokens to every session, about $0.0047 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.