From bc
Analyze normalized requirements and test cases to identify coverage gaps, untested requirements, orphan test cases, and weak coverage.
How this skill is triggered — by the user, by Claude, or both
Slash command
/bc:analyze-coverageThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Identify coverage gaps between requirements and test cases after they have been fetched, normalized, and cross-linked.
Identify coverage gaps between requirements and test cases after they have been fetched, normalized, and cross-linked.
After artifacts have been fetched and normalized. This is the core analysis skill invoked by the coverage agent.
This skill requires normalized requirements AND test cases. If you were handed zero test cases, do NOT report "0% coverage" as a finding — a zero-test-case input almost always means the fetch was skipped or failed. Stop and tell the calling agent to verify the test-case fetch (empty result vs. provider error). Only compute true 0% coverage when the caller has confirmed the scope genuinely has no test cases.
Requirements where linked_ids is empty — no test cases are linked to them.
Test cases where linked_ids is empty — no requirements are linked to them.
Requirements that have linked test cases, but the test cases do not meaningfully exercise the requirement's acceptance criteria. Examples:
linked_ids → untestedlinked_ids → orphanFor large scopes the coverage agent drives this as a per-feature loop, invoking this skill once per feature so all artifacts never sit in context at once. Whether called per-feature or all at once, apply the same logic:
For each requirement that has at least one linked test case:
Important: Weak coverage findings require human judgment. Label them clearly as suggestions, not definitive gaps.
Compute these for every analysis:
(tested / total) * 100Produce a human-readable report:
# Coverage Analysis Report
## Summary
- **Scope**: [what was analyzed]
- **Requirements**: [total] ([tested] tested, [untested] untested)
- **Test cases**: [total] ([linked] linked, [orphans] orphans)
- **Coverage**: [percentage]%
## Feature Coverage Breakdown
| Feature | Requirements | Tested | Untested | Coverage |
|---------|-------------|--------|----------|----------|
| [name] | [n] | [n] | [n] | [n]% |
## Untested Requirements
### [Feature Name]
- **CWA-REQ-XX**: "[title]" — [brief note on what is untested]
## Orphan Test Cases
- **TC-XXXX**: "[title]" — no linked requirement found
## Weak Coverage
- **CWA-REQ-XX**: "[title]" — linked to TC-XXXX but test only covers [aspect]; requirement also specifies [missing aspect]
## Recommendations
1. [Highest priority: safety-critical untested requirements]
2. [Feature areas with lowest coverage %]
3. [Orphan test cases that may need linking or removal]
| Severity | Criteria |
|---|---|
| HIGH | Safety-critical or core functionality requirement with zero test coverage |
| MEDIUM | Non-critical requirement with zero coverage, or orphan test case |
| LOW | Weak coverage — test exists but is superficial |
Provides a checklist for code reviews covering functionality, security, performance, maintainability, tests, and quality. Use for pull requests, audits, team standards, and developer training.
npx claudepluginhub omar-hegazy-integrant/buddy-council-plugin --plugin bc