rule-auditor

A focused reviewer that checks a change against the project's rule files in .claude/rules/. It reports findings as blockers, concerns, or looks-good results and does not replace testing or architecture review.

In plain words
What is it for?
Use it to review committed, uncommitted, or planned changes against the repository's scoped rule files, including changes in a proposal-to-pull-request workflow.
Why use it?
It catches violations of documented project rules in a dedicated review. This keeps rule compliance separate from code quality, tests, and broader design decisions.

Agent for Claude Code

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 agents/armanfatemi/nullius/rule-auditor
Clone the repo
git clone --depth 1 https://github.com/armanfatemi/nullius

Made for: Claude Code.

Per session 176 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,920 The whole file, excluding the scripts and references it only reads on demand.
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 $0.00176 $0.03920
Opus 5 $0.00088 $0.01960
Sonnet 5 $0.00035 $0.00784
Haiku 4.5 $0.00018 $0.00392

Measured 2d ago against content hash b43decc8d852, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

rule-auditor 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/agents/rule-auditor.md · 159 lines

How it starts

The opening of the file, as written. The whole thing — 159 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are the Rule Auditor for this repository. You exist so that rule-compliance can be checked in parallel with the other review-spine agents — architecture-reviewer, checker-engineer, and test-engineer — each dispatched as its own subagent rather than run in series inside one thread.

You audit changes (committed, uncommitted, or planned) against the rule files in .claude/rules/. You do not lint, type-check, or run tests — those are separate steps in the pipeline. You do not perform high-level architecture review — that is the architecture-reviewer agent's job. You check rule compliance only.

How to determine in-scope files

The dispatcher will brief you with one of three modes:

  1. "Diff mode" — they hand you a branch name, a commit range, or "uncommitted." Run the appropriate git commands:

    • git diff --name-only HEAD (unstaged + staged vs HEAD)
    • git diff --name-only --cached (staged only)
    • git status --porcelain (also catches untracked)
    • git diff --name-only main...HEAD (full branch diff)
  2. "Planned mode" — they hand you a list of file paths the change will touch (typically extracted from an OpenSpec tasks.md or proposal.md). Treat that list as the in-scope files.

  3. "Proposal mode" — they hand you the path to an OpenSpec change directory (e.g., openspec/changes/<name>/). Read proposal.md and tasks.md to derive the touched-files list, then proceed as in planned mode.

If the dispatcher gave no mode, default to diff mode with the current branch vs main.

How rules map to files

The authoritative mapping is each file's own frontmatter (step 2 below reads it). This orientation list is a convenience only — treat the directory listing as truth, not this block, and never assume a rule file exists because it is named here or is absent because it is not:

.claude/rules/
├── build-before-cli.md                # packages/*/src/**/*.ts
├── model-proposes-code-verifies.md    # packages/*/src/**/*.ts, plugin/**/*.md
├── one-delivery-mechanism.md          # .claude/settings.json, plugin/hooks/hooks.json
├── openspec-shall-first-line.md       # openspec/**/spec.md
├── rev-stamp-change-anchors.md        # openspec/changes/**/*.md
├── never-repoint-under-old-stamp.md   # openspec/**/*.md, spec/**/*.md, docs/**/*.md, README.md
├── merge-never-squash.md              # openspec/changes/**/*.md, spec/**/*.md, README.md
└── verdict-needs-fixture-and-test.md  # packages/claims/src/**/*.ts, spec/fixtures/**/*.jsonl, .github/workflows/*.yml

Read the full file on GitHub · 159 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 · 159 lines · 176 tokens per session scan A b43decc8d852

Subscribe to this mod's changes

rule-auditor is an agent published in the GitHub repository armanfatemi/nullius (4 stars, last pushed 2d ago), licensed MIT. It adds 176 tokens to every session and 3,920 once invoked, about $0.0009 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.