@spardutti/claude-skills 2.37.1 → 2.38.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +1 -0
- package/hooks/skill-application-gate.sh +2 -0
- package/hooks/skill-gate.sh +2 -0
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -135,6 +135,7 @@ Portable slash commands installed to `.claude/commands/`. Some orchestrate paral
|
|
|
135
135
|
| `/discover` | Find the right problem before deciding what to build — diverges first: generates competing framings through blind subagents under forced constraints, stress-tests the winner with a blind critic, and refuses to converge until every open question is answered or deferred, then drafts scope, non-goals, edge cases and success criteria for you to correct. Run before `/plan-feature` |
|
|
136
136
|
| `/plan-feature` | Integration-first feature planning, then building — 3 parallel subagents scan for reusable code, patterns and touch points, grounded clarifying questions follow, and on your go the same agent builds the plan |
|
|
137
137
|
| `/refactor` | Detect size / complexity / duplication / coupling issues via 4 parallel subagents, then refactor |
|
|
138
|
+
| `/optimize` | Make one measured job faster without changing what it proves — fixes a benchmark and a result that must not move, sends 4 read-only subagents to find suspects, then changes one thing per round with a prediction written first, keeping only what beats the noise. Profiles after every wrong prediction, reruns before judging a result that flipped, and ends by writing the cause as a skill rule for you |
|
|
138
139
|
| `/deep-review` | Multi-agent deep code review — 5 parallel subagents catch guard bypasses, lost async state, wrong-table queries, dead references, protocol violations |
|
|
139
140
|
| `/test-review` | Write-then-verify test review — scopes to the diff, runs red-green + mutation gates, then an isolated read-only subagent proves each test would catch a regression instead of rubber-stamping it |
|
|
140
141
|
| `/handoff` | Write a handoff doc so a fresh session picks the work up — captures what auto-compaction loses: decisions made and rejected, dead ends already ruled out, why the tree is dirty, and the one next step |
|
|
@@ -43,6 +43,8 @@ if [ "$TOOL" != "Bash" ]; then
|
|
|
43
43
|
# The settings file that configures the escape hatch cannot sit behind the
|
|
44
44
|
# gate, or a denied ack has no way to be un-denied.
|
|
45
45
|
*/.claude/settings.json|*/.claude/settings.local.json) exit 0 ;;
|
|
46
|
+
# /optimize's benchmark lives outside the repo so no round can edit it; it is not project code.
|
|
47
|
+
/tmp/claude-optimize-*) exit 0 ;;
|
|
46
48
|
*.*) printf '%s' "$TARGET" | grep -qiE ".($PROSE_EXT)$" && exit 0 ;;
|
|
47
49
|
esac
|
|
48
50
|
fi
|
package/hooks/skill-gate.sh
CHANGED
|
@@ -61,6 +61,8 @@ if [ "$TOOL" != "Bash" ]; then
|
|
|
61
61
|
# matters: a skill claiming **/* gated it, and nothing could re-open it.
|
|
62
62
|
.claude/settings.json|.claude/settings.local.json) exit 0 ;;
|
|
63
63
|
*/.claude/settings.json|*/.claude/settings.local.json) exit 0 ;;
|
|
64
|
+
# /optimize's benchmark lives outside the repo so no round can edit it; it is not project code.
|
|
65
|
+
/tmp/claude-optimize-*) exit 0 ;;
|
|
64
66
|
*.*) printf '%s' "$TARGET" | grep -qiE ".($PROSE_EXT)$" && exit 0 ;;
|
|
65
67
|
esac
|
|
66
68
|
TARGETS="$TARGET"
|