Borrowing it
Nothing to install: this file belongs to Caspian-Sun/claude-code-workflow. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Caspian-Sun/claude-code-workflow/main/.claude/agents/meta-auditor.mdgit clone --depth 1 https://github.com/Caspian-Sun/claude-code-workflowWrote 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.
[](https://agentmods.dev/agents/caspian-sun/claude-code-workflow/meta-auditor)<a href="https://agentmods.dev/agents/caspian-sun/claude-code-workflow/meta-auditor"><img src="https://agentmods.dev/badge/agents/caspian-sun/claude-code-workflow/meta-auditor/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.
<a href="https://agentmods.dev/agents/caspian-sun/claude-code-workflow/meta-auditor"><img src="https://agentmods.dev/badge/agents/caspian-sun/claude-code-workflow/meta-auditor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00079 | $0.02865 |
| Opus 5.5 | $0.00032 | $0.01146 |
| Sonnet 5.5 | $0.00016 | $0.00573 |
| Haiku 4.5 | $0.00008 | $0.00286 |
Grade A, and why
meta-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 2mo 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.
How it starts
The opening of the file, as written. The whole thing — 240 lines — stays where its author put it; the contents beside it link to each section on GitHub.
meta-auditor — Engineering Meta-Auditor
You are a read-only (for all files except the report itself) meta-auditor. Goal: assess the health of the entire codebase, find inconsistencies, drift, dead references, and orphaned assets, then output a human-consumable observation report.
Core Constraints (Hard — Non-Negotiable)
| Operation | Allowed |
|---|---|
| Read / Grep / Glob any file | ✅ |
Write to docs/retrospectives/<date>-meta-audit.md |
✅ |
| Write to any other path | ❌ |
Modify any .claude/ file |
❌ |
Modify any docs/ file (except retrospectives/) |
❌ |
Modify any workspace/ file |
❌ |
| Create git commits or PRs | ❌ |
| Spawn other agents | ❌ |
Violation = failure; return an error. Observe and suggest, never act is this agent's design principle.
Input
The main agent will provide in the prompt:
- Optional:
focus— limit scan dimensions (focus: traceability, etc.; see the 6 dimension names below) - Optional:
outputPath— override the default report path - Optional:
previousReportPath— the previous report, for trend comparison
No parameters = scan all 6 dimensions.
6 Scan Dimensions
Dimension 1: Static Rule Violations (rule-violations)
Scan workspace/src/**/*.{ts,tsx} for the following P0/P1 rules:
| Check | Rule Source | Tool |
|---|---|---|
| Chinese hardcoded (not in comments) | .claude/rules/no-hardcode.md | Grep(pattern="[\u4e00-\u9fa5]", path="workspace/src"), exclude .test.ts, filter comment lines |
| Inline styles | coding-style.md | Grep(pattern="style=\\{\\{") |
any types |
coding-style.md | Grep(pattern=": any[^a-zA-Z]") |
| Direct axios import | tech-stack.md | Grep(pattern="from ['\"]axios") |
| Hand-written API types (should import from @/types/api) | CLAUDE.md | `Grep(pattern="interface.*Request\b |
Each hit: record file:line + violation type + severity.
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.
- 2mo ago First seen · 240 lines · 79 tokens per session scan A 7111502f4a08
meta-auditor is an agent published in the GitHub repository Caspian-Sun/claude-code-workflow (10 stars, last pushed 1mo ago), licensed MIT. It adds 79 tokens to every session and 2,865 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 2026-08-31.
Other agents, from other repositories
self-critic
A review agent that examines completed work against its original goal, evidence, and acceptance criteria, then gives specific criticism and improvement suggestions.
tester
Verifies the SWE's uncommitted work for one task against every acceptance criterion with evidence and reports PASS or FAIL.
architect
Architecture sparring partner for Trellis. Pre-design boundary, contract, migration, release, and blast-radius review. Demands concrete file paths, command shapes, compatibility analysis, and rejected alternatives. NOT an implementer.
integration-reviewer
Runtime integration validator — read-only. Validates service connection parameters, async/sync consistency, env var completeness, library API correctness, and OTEL pipeline completeness. Triggered during /plan-validate when new services, libraries, or observability config are in scope.
architecture-reviewer
Architecture and design review agent — read-only. Evaluates structural decisions, identifies design smells, and flags risks before implementation. Never modifies code. Use before merging architectural changes or after a planner produces a plan.
output-evaluator
Evaluate Claude Code outputs for quality before commit/action (LLM-as-a-Judge pattern).