agentleFS
hegeldevhegeldev/hegel-skill/skills/hegel-review/SKILL.md

hegel-review

Skill90 starsChanged 2 months agoMITView on GitHub

Review property-based tests for the failure modes that make them weak or misleading. Use after writing property-based tests with hegel (or proptest, quickcheck, fast-check, jqwik, etc.), when asked to review property-based tests, or before shipping a change to a PBT suite.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Check every test against this list.

    line 12

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Reviewing property-based tests

Check every test against this list. When a check fails, quote the offending code and give the smallest change that fixes it.

  1. Narrowed generators: drawn domains trimmed to dodge failures — ranges shrunk, characters filtered, sizes capped in the domain rather than at materialization, or a whole value class (subnormals, tiny magnitudes, exact bounds) excluded after it failed there, with or without an explanatory comment. The domain should be as wide as the contract.
  2. Fixed inputs in property clothing: a "property" that only ever sees one value or a few hand-picked cases.
  3. Missing direction: a parser or validator tested only on valid input; or invalid input checked only for "doesn't crash" when rejection should be asserted.
  4. More than one property per test: assertions about unrelated contracts sharing one test and one failure signal.
  5. Evidence-free properties: asserting behavior nothing documents or implies, or restating the implementation instead of a contract.
  6. Tolerance hacks: an epsilon widened until the test passes; approximate equality where the contract is exact. Equality, ordering, and hashing laws, exact roundtrips, and endpoint or identity claims (ratio 0 or 1 returning an endpoint) are exact contracts; a failed check on one dismissed as "a precision limitation", or a tolerance added after a failure was observed with no ledger note grounding it in the contract, is this hack applied during triage.
  7. Weakened oracles: asserting only is_ok(), only a length, or only "doesn't panic" where the actual value could be checked cheaply.
  8. should_panic or ignored tests: tests that pass while the bug is present, or are excluded from the default run.
  9. Config parameters fixed: constructor and configuration knobs pinned to defaults instead of generated.
  10. Untested variants: one property instantiated for one type or variant when the contract covers several; or the same property copy-pasted with drift.
  11. Silent-triage residue: commented-out asserts, deleted failing tests, TODOs hiding a red test.
  12. Resource bounds in the wrong place: recursion depth or collection size limited in the drawn domain instead of at materialization.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.