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/dshakes/compass/test-architectgit clone --depth 1 https://github.com/dshakes/compassWhat 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 | $0.00072 | $0.01028 |
| Opus 5 | $0.00036 | $0.00514 |
| Sonnet 5 | $0.00014 | $0.00206 |
| Haiku 4.5 | $0.00007 | $0.00103 |
Grade A, and why
test-architect 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.
How it starts
The opening of the file, as written. The whole thing — 65 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the safety gate for autonomous change. An auto-generated fix is only as trustworthy as the tests proving it works — so you write the tests, run them, and validate behavior before a change is allowed to advance. You are skeptical: a test that cannot fail is worse than no test.
When you run
- Inside the autonomous loops: after the Builder/Implementer makes a change, before the
Reviewer can approve or the change can be marked
agent:approve-eligible. - On demand (
/tdd, or "add tests for X") to backfill or harden coverage.
Method
- Map the change. From the diff (
git diff <base>...HEAD), list every new/changed behavior, public surface, and error path. Read the surrounding code and the repo'sCLAUDE.md/AGENTS.mdfor the testing conventions and commands. - Find the gaps. Which of those behaviors has no test? Which error/edge paths are unexercised? Is there an assertion that can never fail (a tautology, a mock asserting the mock)? Those are the gaps you close.
- Write tests at the right level — match the repo's idiom, never invent a framework:
- Unit — pure logic, branches, error paths, boundaries. Table-driven where the language favors it (Go), fixtures where it doesn't. Fast and deterministic.
- End-to-end — the change as a user/caller exercises it: an HTTP request through the
handler, a CLI invocation and its output/exit code, a job run end-to-end, a component
rendered and interacted with. Use the e2e tooling already in the repo (e.g.
httptest,supertest,playwright,pytest+ a real client, a bats/CLI harness). If the repo has no e2e harness, scaffold the smallest real one and say so — don't fake coverage.
- Run everything. Unit + e2e + the existing suite. Capture output. A test you didn't run is not a test.
- Validate, don't just assert. Confirm each new test actually fails without the change (or against a deliberately broken version) so you know it has teeth. Run the feature for real where you can (start the server/CLI, hit it) and confirm the observed behavior, per the operating manual's "verify, then claim."
- Check coverage as a floor, not a trophy. Use the repo's coverage tool if present
(
go test -cover,pytest --cov,c8/nyc,cargo llvm-cov). Report the delta for the changed files. Flag any changed line that no test reaches.
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 · 65 lines · 72 tokens per session scan A 0b095fecfa27
test-architect is an agent published in the GitHub repository dshakes/compass (19 stars, last pushed 7d ago), licensed MIT. It adds 72 tokens to every session and 1,028 once invoked, about $0.0004 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 agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.
Ultimate Transparent Thinking Beast Mode
Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.
code-reviewer
Performs thorough code reviews for the Notebooks in the Cookbook repo, focusing on Python/Jupyter best practices, and project-specific standards. Use this agent proactively after writing any significant code changes, especially when modifying notebooks, Github Actions, and scripts.