From code-index
Debug bugs and assess change safety using dependency tracing and impact analysis
How this skill is triggered — by the user, by Claude, or both
Slash command
/code-index:investigateThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Trace bugs to their root cause and assess the safety of proposed changes.
Trace bugs to their root cause and assess the safety of proposed changes.
search("TypeError in payment processing")search("processPayment") for full contexttrace_dependencies(target="processPayment", dependents=true)diff_analysis(diff_spec="HEAD~5..HEAD")Trace what depends on the target:
trace_dependencies(target="UserModel", dependents=true)
Review depth groups:
Assess risk level (weighted formula: dependents + processes x 8):
| Risk | Score | Action |
|---|---|---|
| CRITICAL | >= 40 | Phased rollout, comprehensive testing |
| HIGH | >= 20 | Thorough review, dedicated test pass |
| MEDIUM | >= 8 | Standard review, verify affected paths |
| LOW | < 8 | Normal review process |
Check both directions for major refactors:
trace_dependencies(target="UserModel", dependents=true, dependencies=true)
| Pattern | Strategy |
|---|---|
| Null reference | search for the field, check assignments via context |
| Wrong output | Follow process trace through data transforms |
| Import error | Check IMPORTS edges in symbol context |
| Race condition | trace_dependencies to find concurrent accessors |
| Regression | diff_analysis to correlate with recent changes |
explore_structure to see if a change crosses module boundariesCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub nnourr/code-index