coverage-gap-finder
Skill1 starsChanged 5 days agoMITView on GitHub
Run the project's coverage tool and identify the highest-value untested paths — error branches, edge cases, and critical modules — then propose specific test cases for each gap. Use when you have a coverage report but don't know where new tests will pay off most.
4 other projects use a version of this.
Adopted by
Most recently taken up by
- Sagargupta16/agent-recipestheir own version, first seen Nov 1, 2026
- mrenigmatic19/claude_skillstheir own version, first seen Oct 8, 2026
- SkillMedev/qa-test-automationtheir own version, first seen Oct 5, 2026
- srikanth-manivannan/claude-agent-factorytheir own version, first seen Jun 26, 2026
Other versions4 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 4
- Sagargupta16/agent-recipes★ 4same name
- srikanth-manivannan/claude-agent-factory★ 1same name
- mrenigmatic19/claude_skills★ 0same name
- SkillMedev/qa-test-automation★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository163 other files it gives its agents.
- agents.mdAGENTS.md
- 3d-asset-handoff-sheetSkill
- accessibility-regression-auditorSkill
- adr-writerSkill
- agent-memory-designerSkill
- agent-trajectory-evaluatorSkill
- alerting-rules-tunerSkill
- analysis-memo-writerSkill
- annotation-disagreement-ledgerSkill
- annotation-task-spec-builderSkill
- api-deprecation-plannerSkill
- api-error-contract-designerSkill
- architecture-diagram-generatorSkill
- asset-metadata-review-ledgerSkill
- auth-flow-reviewerSkill
- background-job-reliability-auditorSkill
- branch-rebaserSkill
- brand-voice-profilerSkill
- bundle-analyzerSkill
- cache-policy-designerSkill
- canary-release-plannerSkill
- caption-correction-ledgerSkill
- changelog-from-prsSkill
- chart-chooserSkill
- chunking-strategy-optimizerSkill
- circular-dependency-breakerSkill
- claude-settings-auditorSkill
- code-example-verifierSkill
- cold-email-deliverability-auditorSkill
- cold-start-optimizerSkill
- commit-splitterSkill
- competitor-teardownSkill
- component-spec-writerSkill
- connection-pool-tunerSkill
- content-repurposerSkill
- contract-obligation-registerSkill
- contract-test-designerSkill
- conventional-commitsSkill
- cors-configuratorSkill
- crm-export-auditorSkill
- dashboard-designerSkill
- data-retention-auditorSkill
- dataset-first-lookSkill
- dead-code-finderSkill
- deadlock-diagnoserSkill
- decision-log-reconcilerSkill
- dependency-auditSkill
- dependency-upgrade-plannerSkill
- design-brief-writerSkill
- design-critique-checklistSkill
- design-token-extractorSkill
- devcontainer-designerSkill
- distributed-tracing-instrumenterSkill
- dockerfile-optimizerSkill
- email-sequence-drafterSkill
- embedding-index-tunerSkill
- embedding-set-inspectorSkill
- entity-match-candidate-ledgerSkill
- exit-ticket-builderSkill
- extract-moduleSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
1 rule. 0 are common, 1 is distinctive, and 0 are very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
1
Distinctive
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
- Distinctive: line 21
Detect the test stack and coverage tool: Do not guess — inspect the project:
“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Turn a raw coverage report into a ranked, actionable plan. This skill runs the project's existing coverage tool, reads the per-line and per-branch data, and surfaces the gaps that actually matter — uncovered error handlers, unguarded edge cases, and critical modules with thin coverage — rather than nudging an arbitrary percentage upward. For each gap it proposes concrete, named test cases you can hand straight to a scaffolder. The goal is risk reduction per test written, not a green 100% badge.
When to use this skill
- You have (or can generate) a coverage report but don't know which untested lines are worth testing first.
- A module is business-critical and you want to confirm its error paths and edge cases are exercised.
- You're triaging tech debt and need a prioritized list of test gaps instead of a wall of red lines.
[!NOTE] Line coverage measures executed lines, not correct behavior. A function can be 100% covered by a test that asserts nothing. Treat coverage as a map of blind spots, not a quality score — prioritize gaps by blast radius, then verify the new tests actually assert something.
Instructions
- Detect the test stack and coverage tool. Do not guess — inspect the project:
- JS/TS:
package.jsonforvitest --coverage/jest --coverage(look forcoveragescripts or ac8/nyc/@vitest/coverage-v8dep). - Python:
pytest --cov(pytest-cov),coverage run, config inpyproject.toml/.coveragerc. - Go:
go test -coverprofile. Java: JaCoCo. Match whatever already runs in CI.
- JS/TS:
- Generate machine-readable coverage. Run the tool to emit a structured report —
--coverage --coverage.reporter=json(Vitest),--cov --cov-report=json(pytest), or-coverprofile=cover.out(Go). Parse the JSON/profile, not the pretty terminal table; you need per-file branch and line data. (For Vitest,--coverage.reporter=jsonwritescoverage/coverage-final.jsonwith per-file branch and line data; the unrelated--reporter=jsonis a test-result reporter — pass/fail/timing — and won't produce coverage.) - Rank gaps by value, not size. For every uncovered region, weight it by signals — uncovered
catch/except/error returns,if/switchbranches with no covered alternative, input validation, and modules central to the app (auth, payments, parsing, persistence). Down-rank generated code, trivial getters, and config glue. A 60%-covered payment module outranks a 40%-covered logging helper. - Locate the exact untested paths. For each top gap, read the source and name the specific uncovered branch (file, function, line range) and why it's untested — e.g. "the
429retry branch is never hit because no test injects a rate-limit response." - Propose concrete test cases. For each gap, write one bullet per missing case stating the input/condition and the expected behavior — not "add tests for error handling," but "call
withRetrywith a stubbed client that throwsRateLimitErrortwice then succeeds; assert it retries and returns the value." Hand these totest-scaffolderto generate. - Verify the baseline and report. Re-run coverage to confirm the numbers you're quoting are current, then output a prioritized gap list (highest value first) with file, current coverage, the risk, and the proposed cases. Flag any module you couldn't analyze (e.g. excluded from the report, untestable without a fixture).
[!WARNING] Never chase 100%. Forcing coverage on glue code, framework boilerplate, or unreachable defensive branches produces brittle tests that assert nothing and slow the suite. Stop when the remaining gaps are low-risk — and say so explicitly in the report.
Examples
Running coverage on a Vitest project and triaging the JSON report:
npx vitest run --coverage --coverage.reporter=json
# reads coverage/coverage-final.jsonOutput — a prioritized gap list, not a flat percentage:
Coverage: 78% lines / 64% branches (412/640 branches)
HIGH VALUE
1. src/payments/charge.ts — 71% lines, 50% branches
Risk: the declined-card and idempotency-key-reuse branches are never exercised.
Propose:
- charge() with a gateway stub returning `card_declined` → asserts no DB write, throws PaymentError
- charge() called twice with the same idempotency key → asserts the second call returns the first result, no double charge
2. src/auth/verifyToken.ts — 83% lines, 60% branches
Risk: expired-token and malformed-signature paths uncovered.
Propose:
- verifyToken() with a token whose `exp` is in the past → asserts throws TokenExpiredError
- verifyToken() with a tampered signature → asserts throws InvalidSignatureError
MEDIUM VALUE
3. src/parse/csv.ts — 88% lines
Risk: quoted-field-with-embedded-comma branch untested.
Propose:
- parseCsv('"a,b",c') → asserts two fields ["a,b", "c"]
SKIP (low value)
- src/logger.ts (45%) — thin wrapper over console; defensive branches only.
- src/generated/*.ts — codegen output, exclude from coverage instead.Hand the HIGH and MEDIUM cases to test-scaffolder, then re-run coverage to confirm the branches now flip green.
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
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.

