agentleFS
Sign inStart free
aezizhuaezizhu/hyperworkflows/skills/oracle-forging/SKILL.md

oracle-forging

Skill0 starsChanged 4 months agoMITView on GitHub

Use when a unit of work has no executable acceptance (no test, no verifiable output) and you need to construct one - golden files, property tests, metamorphic relations, snapshot tests - or prove it infeasible.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    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:

    Test-only changes. Restructuring production code "for testability" changes the thing being tested.

    line 17
  2. Distinctive:

    Every forged oracle is itself verified by an independent verifier run before the unit leaves the grey queue.

    line 19

≈ “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

Oracle Forging

A missing oracle is a work item, not a label. Forge executable acceptance in this preference order — strongest first:

  1. Exact golden file — deterministic output? Capture it: diff <(cmd) fixtures/golden.txt; exit code 0. Strongest oracle; brittle only if the output legitimately varies.
  2. Property-based test — output varies but properties hold (sorted, idempotent, round-trips, conserves length/sum). Encode the property, generate inputs.
  3. Metamorphic relation — no absolute expectation, but relations across runs hold: f(x) == f(shuffle(x)), count(filter(p)) <= count(all), decode(encode(x)) == x.
  4. Snapshot test — last resort: freeze current behavior to detect change, not correctness. Label it as change-detection so nobody mistakes it for a correctness oracle.

Rules:

  • The forged acceptance must PASS on current code. If it fails, you found a defect — report it as a finding, not an oracle.
  • Test-only changes. Restructuring production code "for testability" changes the thing being tested.
  • Declare infeasible only with a concrete reason a human can evaluate ("output is subjective visual layout", "requires NDA hardware"). "Hard to test" is not infeasible.
  • Every forged oracle is itself verified by an independent verifier run before the unit leaves the grey queue.

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.

Check your setup, free