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.
npx agentmods add commands/fotoetienne/gru/auditgit clone --depth 1 https://github.com/fotoetienne/gruWhat 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 | $0.00016 | $0.01620 |
| Opus 5 | $0.00008 | $0.00810 |
| Sonnet 5 | $0.00003 | $0.00324 |
| Haiku 4.5 | $0.00002 | $0.00162 |
Grade A, and why
audit 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 3d 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 — 139 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Run a Code Quality Improvement Plan (CQIP) — a structured codebase audit that identifies test gaps, dead code, duplication, testability issues, and code complexity, then files actionable GitHub issues for all findings.
No source code is modified during this process.
Instructions:
Phase 0: Detect Language & Check for Existing CQIP Issues
-
Detect the project language by checking for
Cargo.toml(Rust),package.json(JS/TS),pyproject.toml/setup.py(Python),build.gradle(Java/Kotlin), etc. Adapt the audit instructions in Phase 1 to use language-appropriate constructs (e.g.,#[test]for Rust,describe/itfor JS,def test_for Python). -
Check for existing open CQIP issues:
gh issue list --label cqip --state open --json number,titleIf open CQIP issues exist, note them. In Phase 4, skip filing issues that duplicate existing open ones.
Phase 1: Audit (parallel agents)
Spawn two auditor agents in parallel using the Agent tool. Each agent should write its findings to a temp file, then send a summary message back.
Agent 1: Test Auditor
Read every test module and test function across the codebase. For each:
- Classify each test as high/medium/low value
- Identify tests to delete: trivial tests, tests that test mocks not behavior, tests with no assertions
- Identify tests testing the wrong thing: tests that re-implement production logic locally, tests that inline production format strings
- Identify dead code kept alive by tests: production code with zero non-test callers
- Identify critical missing coverage: untested code paths most likely to break in production (error handling, state transitions, async loops, timeout machinery)
- Identify tests to improve: tests with no assertions, tests using real CLI tools instead of mocks
Report findings with specific file paths and line numbers, grouped by module.
Agent 2: Architecture Auditor
Read all source files and audit for:
- Testability: Where interfaces/traits should be added for dependency injection, functions mixing business logic with I/O, global state blocking parallel tests
- Long functions: Functions >80 lines with multiple responsibilities — list file:line, length, suggested splits
- Duplication: Repeated patterns across modules that should be extracted into helpers
- Dead code: Unused items, unused public API surface, commented-out code
- API surface: Overly broad visibility (e.g.,
pubitems that should bepub(crate)in a binary crate)
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.
- 3d ago First seen · 139 lines · 16 tokens per session scan A 79110107f24f
audit is a command published in the GitHub repository fotoetienne/gru (11 stars, last pushed 26d ago), licensed Apache-2.0. It adds 16 tokens to every session and 1,620 once invoked, about $0.0001 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 2026-08-30.
Other commands, from other repositories
roadmap-suggest
Mine the repo for new roadmap items — or, on a fresh project, bootstrap the board from scratch.
roadmap-sequence
Report or assemble a Sideboard sequence for the given issue numbers (targeted modes).
roadmap-cleanup
Scan the roadmap for issues that are already done or redundant, and propose closing/merging them.
roadmap
Open the Sideboard roadmap board in the preview pane.
speckit-implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.
speckit-specify
Create or update the feature specification from a natural language feature description.