doctor: make hook sprawl + duplicate checks matcher-aware#77
Merged
Conversation
Dogfooding on a real settings.json exposed two false positives in the Stream-B hook-safety checks: - sprawl counted ALL hooks per event, but a matcher-scoped hook (e.g. bound to mcp__cal__create) does not fork a process on unrelated calls. Severity now keys on always-fire hooks (matcher "" or "*") — the real per-call cost — so 12 PreToolUse hooks each scoped to a distinct tool no longer FAIL. - duplicate detection keyed on (event, command) only, so the same guard under different matchers was wrongly flagged. The dedup key now includes the matcher; only true copy-paste (same matcher + command) is a duplicate. Net: on a clean-but-matcher-heavy config, doctor now reports "All critical checks passed" instead of a spurious FAIL + duplicate warnings. The inventory SCAN line still shows totals. TDD: added matcher-discriminating tests (scoped≠sprawl, different-matcher≠dup). Unit + contract + arch gates green. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
📝 WalkthroughWalkthroughHook sprawl and duplicate detection in ChangesHook analysis matcher-aware refactor
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@internal/hooks/analyze.go`:
- Around line 302-322: Canonicalize matcher aliases before building the dedup
key: in the loop over entries (the one that creates key{matcher, command} and
updates cmdCount/cmdOrder), normalize e.Matcher to a canonical value (e.g.,
treat "" and "*" the same by mapping "" -> "*" or vice‑versa) and use that
normalized value when constructing key; no other structural changes to key,
cmdCount, or cmdOrder are needed. Also add a regression test that constructs
entries with the same command and event but mixed matchers ("" and "*") and
asserts that a Finding with Code "hook-duplicate" is produced, ensuring the new
normalization catches the duplicate.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Pro
Run ID: 0e9c09ea-d26b-4dbd-bc17-05e1edfb7404
📒 Files selected for processing (2)
internal/hooks/analyze.gointernal/hooks/analyze_test.go
Comment on lines
+302
to
322
| // Exact duplicates: same (matcher, command) appearing >1 times. The same | ||
| // command under DIFFERENT matchers is intentional per-tool protection, | ||
| // not a duplicate, so the matcher is part of the dedup key. | ||
| type key struct{ matcher, command string } | ||
| cmdCount := map[key]int{} | ||
| cmdOrder := []key{} | ||
| for _, e := range entries { | ||
| if _, ok := cmdCount[e.Command]; !ok { | ||
| cmdOrder = append(cmdOrder, e.Command) | ||
| k := key{e.Matcher, e.Command} | ||
| if _, ok := cmdCount[k]; !ok { | ||
| cmdOrder = append(cmdOrder, k) | ||
| } | ||
| cmdCount[e.Command]++ | ||
| cmdCount[k]++ | ||
| } | ||
| for _, cmd := range cmdOrder { | ||
| if cmdCount[cmd] > 1 { | ||
| for _, k := range cmdOrder { | ||
| cmd := k.command | ||
| if cmdCount[k] > 1 { | ||
| out = append(out, Finding{ | ||
| Level: LevelWarn, | ||
| Code: "hook-duplicate", | ||
| Message: fmt.Sprintf("%q appears %d times on %s", cmd, cmdCount[cmd], event), | ||
| Message: fmt.Sprintf("%q appears %d times on %s", cmd, cmdCount[k], event), | ||
| Details: []string{"Identical hooks waste resources. Remove the duplicates."}, |
There was a problem hiding this comment.
Canonicalize matcher aliases before duplicate-key counting.
Line [309] uses raw e.Matcher in the dedup key. That misses true duplicates when one hook uses "" and another uses "*" with the same command/event, even though this codebase already treats both as always-fire semantics.
🔧 Suggested fix
func DuplicateFindings(inv *Inventory) []Finding {
+ canonicalMatcher := func(m string) string {
+ m = strings.TrimSpace(m)
+ if m == "*" {
+ return ""
+ }
+ return m
+ }
+
// Group entries by event.
byEvent := map[string][]HookEntry{}
@@
type key struct{ matcher, command string }
cmdCount := map[key]int{}
cmdOrder := []key{}
for _, e := range entries {
- k := key{e.Matcher, e.Command}
+ k := key{canonicalMatcher(e.Matcher), e.Command}
if _, ok := cmdCount[k]; !ok {
cmdOrder = append(cmdOrder, k)
}
cmdCount[k]++
}Please also add a regression test for mixed "" + "*" matchers on the same command/event.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@internal/hooks/analyze.go` around lines 302 - 322, Canonicalize matcher
aliases before building the dedup key: in the loop over entries (the one that
creates key{matcher, command} and updates cmdCount/cmdOrder), normalize
e.Matcher to a canonical value (e.g., treat "" and "*" the same by mapping "" ->
"*" or vice‑versa) and use that normalized value when constructing key; no other
structural changes to key, cmdCount, or cmdOrder are needed. Also add a
regression test that constructs entries with the same command and event but
mixed matchers ("" and "*") and asserts that a Finding with Code
"hook-duplicate" is produced, ensuring the new normalization catches the
duplicate.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Dogfooding the Stream-B hook-safety suite (#33-36) on a real
~/.claude/settings.jsonimmediately exposed two false positives — fixing them here.The bug: on a config where guards are scoped to specific tool matchers (e.g.
calendar_guard.pyundergcal_create_event,gcal_delete_event, …), doctor reported a spuriousFAIL: PreToolUse has 12 hooksand fourhook-duplicatewarnings. But:Bashcall, 0 of those 12 fire. Counting them toward per-call sprawl is wrong.The fix:
""/"*") — the real per-call process cost. The inventory SCAN line still shows totals.(matcher, command); only true copy-paste (same matcher + command) flags.Result: a clean-but-matcher-heavy config now reports "All critical checks passed" instead of a false FAIL.
Verification
task test:go:unit+ contract + arch green.Follow-up fix to the doctor suite shipped in #75.
🤖 Generated with Claude Code
Summary by CodeRabbit