AB test-suite-curation
Audit a whole regression suite and prune/restructure it with evidence: per-test coverage fingerprinting, AST near-duplicate clustering, CI-history mining for never-failing and flaky tests, prune decision rules (redundant/obsolete/low-value/keep), smoke/core/extended tiering by risk and defect-detection history, and a defensible "what we deleted and why" record. Deletion is destructive — quarantine and human sign-off are mandatory. Use when: "audit the test suite," "prune redundant tests," "find duplicate tests," "which tests can we delete," "restructure into smoke/core/extended," "is this test pulling its weight," "shrink the regression suite." Not for: Judging whether an individual test is WELL-WRITTEN (smells, assertions) — that is ai-qa-review. Healing one flaky test at runtime — that is test-reliability. Bulk selector regeneration after a UI refactor — that is selector-drift-recovery. Related: ai-qa-review, coverage-analysis, test-reliability, risk-based-testing, qa-project-context.
Audit a whole regression suite and prune/restructure it with evidence: per-test coverage fingerprinting, AST near-duplicate clustering, CI-history mining for…
As a process B 66/100 · Nearly there — weak spots: result and completion, inputs and preconditions
How to improve
- The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 0
✓ No critical or high findings
Files scanned: 8. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
body-longSKILL.md body ≈ 5999 tokens (recommended < 5000); move details to references/ - note
edit-residuethe text marks something as outdated (lines 14, 38, 40, 119, 123, 148): check that old rules are not kept next to new ones — the full check reads the text for contradictions
Process rating: all ten parameters 66/100
- 0Inputs and preconditions. Does not say what the process needs to start
- 40Result and completion. Does not say what the result is
- 50When it triggers. No condition that starts the skill
- 60Tools and files. Uses tools (bash, python) that frontmatter does not declare
- 70Execution cost. Instruction body is 5999 tokens
- 100Steps. 55 steps
- 100Failures and branches. 1 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Running it twice. Mutating operations check current state
- 100Progress reporting. Reports progress
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low 16 top-level sections: this looks like several domains in one skill
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +3Description length 1001: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- +2Single-language instructions
- +5Description quotes 8 example trigger phrases
- +4Description says when NOT to use the skill
- +4Structure: 25 headings
- +3Step-by-step instructions: 55 items
- +4Has examples (1 code blocks)
- +4Reference files are cited in the instructions (7 of 7)
- +1License stated
Quality base 70; lint remarks subtract, signals add up to 100. Result: 85.