GravionLabs/elysion / .agents/skills/implementation
implementation
VerifiedRules for code changes: every change ships with tests, dispatches to <lang>-implementation. Use for implement, write code, add feature, test requests.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 3.3 KB |
| references/format.md | markdown | 9.6 KB |
| references/lint.md | markdown | 1.8 KB |
| references/runner-contract.md | markdown | 2.3 KB |
| references/specs.md | markdown | 1.2 KB |
| references/test-quality.md | markdown | 1.4 KB |
| schema/config.schema.json | config | 6.3 KB |
| schema/implementer.schema.json | config | 949 B |
| scripts/Format-Changed.ps1 | script | 18.8 KB |
| templates/spec.md | markdown | 607 B |
references/test-quality.md
View on GitHub# Good tests
Adapted from `skills/engineering/tdd` in [mattpocock/skills](https://github.com/mattpocock/skills) (MIT License, Matt Pocock), commit `49dd158d1076134a641b33efb035946536778336`. Reworded for this hub.
## Rules
- **Behavior, not internals.** A test calls the public interface (function, endpoint, command, component) and asserts what a caller can observe. It survives a refactor that keeps the behavior. A test that breaks when a private helper is renamed, or that asserts call order or internal state, is coupled to internals.
- **No tautological assertions.** The expected value comes from a literal, a worked example or the spec, never from re-running the logic under test (`expect(f(x)).toBe(f(x))`, or recomputing with the same formula). If the test would still pass with the implementation broken, it proves nothing.
- **Vertical slices.** Write one failing test, then the minimal code to pass it, then the next test. Never write all tests first and then all code: bulk tests lock in imagined behavior and guess at the interface.
- **Refactor in review, not in the loop.** The red → green loop only makes the test pass. Restructuring happens afterwards, with the tests green, and is judged in review.
## Checks before committing
- Each new test names a behavior and would fail if that behavior broke.
- Each expected value can be traced to a literal, an example or the spec.
- No test reaches into private members or asserts on how the result is produced.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |