Use when a test passes but may be checking the wrong signal, mocks its own assumption, snapshots unstable output, or cannot distinguish correct behavior from a broken implementation.
Scanned 9/3/2026
Install to Claude Code
npx -y skills add ShugokiFable/Ultimate-AI-Starter-Bundle --skill test-oracle-quality --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Test Oracle Quality?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/shugokifable-test-oracle-quality)More formats (shields.io, HTML) on the badges page.
---
name: test-oracle-quality
description: Use when a test passes but may be checking the wrong signal, mocks its own assumption, snapshots unstable output, or cannot distinguish correct behavior from a broken implementation.
---
# Test Oracle Quality
## Core rule
A test is valuable only if its **oracle** fails when the user-visible contract is wrong.
Before trusting a green test, name one realistic production defect that should make it fail. If the test would still pass, strengthen the observation boundary. Prefer final state, parsed output, exit status, hashes, or real protocol behavior over "mock was called".
Watch for **false positive** tests: empty loops, skipped assertions, broad exception swallowing, fixtures that reproduce the implementation bug, or mocks that encode an invented API.
For critical regressions, perform a temporary **mutation** or revert of the fix and confirm the test turns red, then restore and rerun green. A test that never demonstrated failure is weak evidence.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!