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 skills add brcampidelli/chimera-agent --skill chimera-prove-the-test-discriminatesgit clone --depth 1 https://github.com/brcampidelli/chimera-agentWrote 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/skills/brcampidelli/chimera-agent/chimera-prove-the-test-discriminates)<a href="https://agentmods.dev/skills/brcampidelli/chimera-agent/chimera-prove-the-test-discriminates"><img src="https://agentmods.dev/badge/skills/brcampidelli/chimera-agent/chimera-prove-the-test-discriminates/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/skills/brcampidelli/chimera-agent/chimera-prove-the-test-discriminates"><img src="https://agentmods.dev/badge/skills/brcampidelli/chimera-agent/chimera-prove-the-test-discriminates.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00035 | $0.01114 |
| Opus 5 | $0.00017 | $0.00557 |
| Sonnet 5 | $0.00007 | $0.00223 |
| Haiku 4.5 | $0.00003 | $0.00111 |
Grade A, and why
chimera-prove-the-test-discriminates 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 11d 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 — 108 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Trigger
You fixed a bug and wrote a test so it cannot come back. The test passes. That is the whole evidence you have, and it is the same evidence you would have if the test asserted nothing about the bug at all.
It applies hardest when the test was written after the fix, because then the only code the test has ever run against is code where the bug is already gone. It applies less to a test written first, red, in a TDD loop — you already watched that one fail, which is the entire point of the loop.
It does not apply to a test for genuinely new behaviour, where there is no "before" to revert to. There, the discriminating check is different: delete the new implementation and confirm the test fails on the absence, not on the bug.
Do
- Identify the exact edit that constitutes the fix — the hunk, not the commit. If the fix is one changed comparison, that comparison is what you are going to undo.
- Undo it.
git stashthe fix, or invert the one line back to its broken form by hand. - Run only the new test:
pytest tests/test_thing.py::test_the_regression -x. - Read the failure. It must be your assertion failing, with a message about the behaviour —
not a
SyntaxError, not a collection error, not anImportErrorfrom a half-reverted file. Those are red for the wrong reason and prove nothing about the test. - Restore the fix (
git stash pop) and confirm the same test now passes. - Put the fact in the test's docstring: what was reverted, and what the failure looked like. The next person to touch this test needs to know it was ever exercised against the bug.
Avoid
Asserting on something true in both worlds. The classic shape:
# WRONG — passes against the buggy code too
result = parse_config(raw)
assert result is not None
assert "timeout" in result
The bug was that timeout came back as the string "30" instead of the int 30. Both assertions
hold either way. What discriminates is the thing that changed:
# RIGHT — fails against the buggy code
assert parse_config(raw)["timeout"] == 30
assert isinstance(parse_config(raw)["timeout"], int)
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.
- 11d ago First seen · 108 lines · 35 tokens per session scan A 90d4bf00eb67
chimera-prove-the-test-discriminates is a skill published in the GitHub repository brcampidelli/chimera-agent (23 stars, last pushed today), licensed Apache-2.0. It adds 35 tokens to every session and 1,114 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 2026-08-30.
Other skills, from other repositories
verify-changes
Verify code changes by running the project's typecheck, build, lint, and targeted tests, then fix and re-run until clean. Use after editing any source file.
deepeval
Use when discussing or working with DeepEval (the python AI evaluation framework).
iterative-refinement
Disciplined, measurable iteration for a substantial refinement or investigation: loop against verifiable pass/fail conditions, fan work out to subagents, and keep the main context lean. Use when improving something measurable over repeated cycles (tuning a metric or detector, refactoring against a regression bar)…
ap-policies
Attach a completion policy gate (shell check or judge-agent rubric) to a session or fan-in group so turn-end only passes when the gate is green, then optionally auto-commit pending human ack. Use when the user says "gate this session on tests passing", "attach a policy", "commit only if tests are green", "judge…
GitHub Issues
Triage GitHub issues, extract the real ask, and connect each issue to the relevant files, tests, or release surface before acting.
Test Driven Development
Drive behavior changes through a failing test, the smallest working implementation, and then cleanup once the behavior is locked.