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 agents/coltonbearden/carrel/test-engineergit clone --depth 1 https://github.com/coltonbearden/carrelWhat 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 | $0.00032 | $0.00233 |
| Opus 5 | $0.00016 | $0.00117 |
| Sonnet 5 | $0.00006 | $0.00047 |
| Haiku 4.5 | $0.00003 | $0.00023 |
Grade A, and why
test-engineer 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 yesterday.
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.
What it actually says
You are the test engineer for the carrel project.
Rules:
- Read
specs/14-fixtures.md,docs/ARCHITECTURE.md§Testing, andCLAUDE.mdfirst. - Fixtures are generated programmatically (
tests/fixtures/generate.py) — idempotent, deterministic where possible, committed outputs. Never hand-craft binary files. - Integration tests drive the real CLI (
uv run carrel ...via subprocess or CliRunner) on real fixtures — no mocking of the filesystem. - Tests that need an optional binary skip with a reason when it's missing; on THIS machine most binaries exist, so verify they actually run here.
- Every claim in your report must come from executed pytest output. Paste the tail.
Completion report format: files touched · commands run + real output tails · coverage gaps you noticed · open issues.
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.
- yesterday First seen · 17 lines · 32 tokens per session scan A a80acb2e29b4
test-engineer is an agent published in the GitHub repository coltonbearden/carrel (1 stars, last pushed 21d ago), licensed MIT. It adds 32 tokens to every session and 233 once invoked, about $0.0002 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.
Other agents, from other repositories
device-review
Mobile checkpoint. Builds a dev client, boots iOS simulator and Android emulator, captures screenshots of changed screens, and asserts performance budgets before QA. Replaces local-review for mobile projects. Trigger on device review, mobile checkpoint, simulator, emulator, visual review of an app, or pre-QA mobile…
maestro-tester
QE E2E testing agent for Flutter apps using Maestro. Converts PM test cases into Maestro YAML flows, ensures TestKeys exist in Flutter code, builds app, runs tests on emulator, reports pass/fail per test case. Use after unit/widget tests pass to verify user journeys.
quality-checker
LLM-as-Judge quality evaluation + non-blocking browser smoke in Phase 5; report-only. Deterministic build/typecheck/lint/test verdicts are @verification-agent's job — this agent's deterministic re-runs are input to the LLM score, not the gate.
test-writer
Write unit, integration, component, and E2E tests for the project's stacks with TDD support.
demo-frame-critic
Reviews a rendered demo video by reading extracted frames and grading each section against its successCriteria — finds errors, empty states, spinners, placeholder data, exposed secrets, clipped text, caption collisions, and pacing problems. Returns a ranked findings list with the stage to re-run for each. Use after a…
geo-qa-verifier
Performs independent QA, regression testing, build verification, contract validation, smoke checks, and final PASS/PASS WITH ISSUES/FAIL reports across GEO Optimizer and GeoReady.