From bc
Analyze normalized requirements and test cases to find contradictions, inconsistencies, and missing alignment. Core analysis engine for the contradiction agent.
How this skill is triggered — by the user, by Claude, or both
Slash command
/bc:detect-contradictionsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Analyze normalized requirements and test cases to find contradictions, inconsistencies, and missing alignment.
Analyze normalized requirements and test cases to find contradictions, inconsistencies, and missing alignment.
After artifacts have been fetched and normalized. This is the core analysis skill invoked by the contradiction agent.
This skill requires BOTH normalized requirements AND test cases as input. If you were handed zero test cases (or zero requirements), do NOT produce a contradiction report — there is nothing valid to compare against. Stop and tell the calling agent the required data is missing so it can re-fetch or report the gap. Never fabricate test cases or treat requirements' linked_ids as a substitute for fetched test cases.
Detect these seven types of issues:
A requirement states X, but a linked test case validates the opposite of X.
Different expected behaviors for the same trigger or condition across linked artifacts.
Two requirements cover overlapping functionality with incompatible constraints.
Test steps or expected results contradict the requirement they claim to verify.
Requirements in different features that create conflicting expectations.
Conflicting expectations about system state or timing.
linked_ids as missing alignmentFor large scopes the contradiction agent drives this as a per-feature loop — it invokes this skill once per feature (intra-feature types 1–4, 6, 7) and once more for the cross-feature pass (type 5), so all artifacts never sit in context at once. Whether called per-feature or all at once, apply the same logic:
If the total artifact count exceeds ~50 items:
Produce a human-readable report structured as:
# Contradiction Analysis Report
## Summary
- **Scope**: [what was analyzed]
- **Requirements analyzed**: [count]
- **Test cases analyzed**: [count]
- **Issues found**: [count by severity]
## Critical Issues
### [CRITICAL] Direct conflict between REQ-85 and TC-1234
**Requirement (CWA-REQ-85)**: "The system shall display patient vitals with a refresh rate of 3 seconds"
**Test Case (TC-1234), Step 3**: Expects "vitals update every 10 seconds"
**Explanation**: The test case validates a 10-second refresh interval, but the requirement specifies 3 seconds. The test would pass for a system that violates the requirement.
**Recommendation**: Update TC-1234 Step 3 to expect a 3-second refresh rate.
## High Issues
...
## Medium Issues
...
## Low Issues
...
## Missing Alignment
### Requirements without test coverage:
- CWA-REQ-92: "System shall support audit logging" — no linked test cases
### Orphan test cases:
- TC-1456: "Verify backup schedule" — no linked requirement
| Severity | Criteria |
|---|---|
| CRITICAL | Direct conflict — test validates the opposite of what requirement states |
| HIGH | Behavioral conflict — test and requirement describe meaningfully different behavior |
| MEDIUM | Scope overlap or cross-feature tension — potential conflict that needs human judgment |
| LOW | Missing alignment — untested requirement or orphan test case |
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