agentleFS
imtiazrayhanimtiazrayhan/agentscamp-library/skills/coverage-gap-finder/SKILL.md

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

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
  1. Distinctive:

    Detect the test stack and coverage tool: Do not guess — inspect the project:

    line 21

“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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

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

  1. Detect the test stack and coverage tool. Do not guess — inspect the project:
    • JS/TS: package.json for vitest --coverage / jest --coverage (look for coverage scripts or a c8/nyc/@vitest/coverage-v8 dep).
    • Python: pytest --cov (pytest-cov), coverage run, config in pyproject.toml/.coveragerc.
    • Go: go test -coverprofile. Java: JaCoCo. Match whatever already runs in CI.
  2. 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=json writes coverage/coverage-final.json with per-file branch and line data; the unrelated --reporter=json is a test-result reporter — pass/fail/timing — and won't produce coverage.)
  3. Rank gaps by value, not size. For every uncovered region, weight it by signals — uncovered catch/except/error returns, if/switch branches 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.
  4. 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 429 retry branch is never hit because no test injects a rate-limit response."
  5. 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 withRetry with a stubbed client that throws RateLimitError twice then succeeds; assert it retries and returns the value." Hand these to test-scaffolder to generate.
  6. 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.json

Output — 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.

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.