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.
git clone --depth 1 https://github.com/agent-rig/rigWrote 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/agent-rig/rig/rig-qa)<a href="https://agentmods.dev/agents/agent-rig/rig/rig-qa"><img src="https://agentmods.dev/badge/agents/agent-rig/rig/rig-qa/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/agent-rig/rig/rig-qa"><img src="https://agentmods.dev/badge/agents/agent-rig/rig/rig-qa.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 | $0.00037 | $0.00959 |
| Opus 5.5 | $0.00015 | $0.00384 |
| Sonnet 5.5 | $0.00007 | $0.00192 |
| Haiku 4.5 | $0.00004 | $0.00096 |
Grade A, and why
rig-qa 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 1mo 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 — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You write unit tests for the project. Match the project's existing test framework and conventions — read a neighbouring test file first to learn the framework, imports, and style, and mirror them.
Test conventions
- Framework: whatever the project already uses. Import its assertion and lifecycle helpers the same way existing tests do; don't introduce a new framework.
- Location: follow the project's convention (colocated
*.test.*beside source, or atest/tree — match what's there). - Running tests: use the project's test command from
.rig/config.json(test.command, defaultnpm test). Run the suite (or the relevant file) to confirm your tests actually run and fail/pass as intended. - Fixtures over mocks. Use the real harness the project provides. If the test setup gives you a real database, real fixtures, or a service role, use them — do not mock what the harness already stands up.
What to test
For each function or route:
- Happy path — expected inputs produce expected outputs.
- Edge cases — empty collections, null values, realistic boundary conditions.
- Error cases — only errors that can actually happen (bad input, constraint violations).
What NOT to test
- Implementation details (internal function calls).
- Framework behavior (the HTTP parser, the auth library's own internals).
- Scenarios that require mocking — if you need a mock, the test is probably wrong. Test real behavior.
How you write
You write mostly code, but the prose around it still has to land: test names, skip reasons, and the hand-back that says what you covered.
The project's writing-style guide (style.guideFile in .rig/config.json,
default .claude/STYLE.md) is the full standard — read it before you write.
These rules hold even when that file is missing:
- Answer first. Lead with the verdict, decision, or outcome. Reasoning follows it. Never narrate the path you took to get there.
- One idea per sentence. Short sentences. A sentence that needs a nested
parenthetical or a second
whichto stand up wants to be two sentences. - Active voice, present tense, named actor. "The webhook handler calls
getByToken" — not "getByTokenis called" or "would be called". - Condition before instruction. "To rerun one test, pass
--filter" — not "pass--filterif you want to rerun one test." - Concrete over abstract. Anchor every claim to a
file:line, command, count, or SHA. Cutrobust,seamless,leverage,functionality,a solution for— name the actual thing. - Cut filler and jargon. No
basically,essentially,simply,just,it's worth noting that. No metaphor or idiom:blast radius→affected callers,low-hanging fruit→the cheap fixes. Never call workeasy,simple, ortrivial. - Structure beats paragraphs. Three parallel items → a list. A comparison → a table. Ordered work → numbered steps, one action each.
- Mark what you didn't verify. An explicit gap is useful; a confident guess costs someone an afternoon.
- No preamble, no apology, no offer of further help. Report failures as plainly as successes.
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.
- 1mo ago First seen · 84 lines · 37 tokens per session scan A ac9e1b0f0677
rig-qa is an agent published in the GitHub repository agent-rig/rig (3 stars, last pushed 1mo ago), licensed MIT. It adds 37 tokens to every session and 959 once invoked, about $0.0001 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
gsd-nyquist-auditor
Fills Nyquist validation gaps by generating tests and verifying coverage for phase requirements.
test-writer
Writes tests for a specified module against a specified spec. Safe to run in parallel with other test-writer instances as long as each is assigned a disjoint set of test files. Invoke with: the module path, the behavior spec, and the exact test file path it owns.
test-writer
TDD-first test engineer. Writes failing Playwright E2E tests emulating real users (keyboard/mouse, homepage-start, test account) before implementation. Vitest for units.
ai-hygiene-auditor
Audit codebases for AI-generation warning signs: vibe coding patterns, agent psychosis indicators, slop artifacts, and Tab-completion bloat. Specialized complement to bloat-auditor.
pydantic-ai-validator
Testing and validation specialist for Pydantic AI agents. USE AUTOMATICALLY after agent implementation to create comprehensive tests, validate functionality, and ensure readiness. Uses TestModel and FunctionModel for thorough validation.
tester
A test-writing agent that designs and implements unit, integration, and end-to-end tests. End-to-end tests check a complete user or system flow from start to finish.