CtrlK
BlogDocsLog inGet started
Tessl Logo

claude-code-hooks

Claude Code hook system for pre/post tool execution. Triggers on: hooks, PreToolUse, PostToolUse, hook script, tool validation, audit logging.

Invalid
This skill can't be scored yet
Validation errors are blocking scoring. Review and fix them to unlock Quality, Impact and Security scores. See what needs fixing →
SKILL.md
Quality
Evals
Security

Evaluation results

95%

25%

Compliance Audit Hook for Claude Code

PostToolUse audit logging hook

Criteria
Without context
With context

Config file location

50%

50%

PostToolUse event

100%

100%

Bash matcher

100%

100%

Command type

100%

100%

Timeout field

0%

100%

stdin read pattern

100%

100%

jq parsing

0%

100%

CLAUDE_PROJECT_DIR usage

100%

100%

Quoted variables

100%

100%

jq error handling

50%

100%

Without context: $0.2125 · 51s · 16 turns · 20 in / 2,633 out tokens

With context: $0.3867 · 1m 15s · 22 turns · 28 in / 4,037 out tokens

100%

20%

Bash Command Safety Guard for Claude Code

PreToolUse blocking hook with security validation

Criteria
Without context
With context

PreToolUse event

100%

100%

Command type and field

100%

100%

stdin read pattern

100%

100%

jq to extract input

0%

100%

Exit code 2 for blocking

100%

100%

Exit code 0 for allowing

100%

100%

Stderr error message

100%

100%

Path traversal check

100%

100%

Quoted variables

100%

100%

jq error handling

0%

100%

Without context: $0.4404 · 1m 46s · 30 turns · 30 in / 6,601 out tokens

With context: $0.4988 · 1m 42s · 26 turns · 717 in / 6,245 out tokens

86%

37%

Session Lifecycle Hooks for Claude Code

Session lifecycle hooks (SessionStart and SessionEnd)

Criteria
Without context
With context

Config file location

25%

50%

SessionStart event

0%

100%

SessionEnd event

0%

100%

No matcher on session events

26%

100%

Command type

100%

100%

CLAUDE_PROJECT_DIR in commands

100%

100%

Timeout set

0%

0%

Quoted variables in scripts

100%

100%

Correct array structure

100%

100%

Without context: $0.3000 · 1m 28s · 15 turns · 17 in / 5,375 out tokens

With context: $0.3095 · 1m 2s · 21 turns · 25 in / 3,635 out tokens

Repository
NeverSight/skills_feed
Evaluated
Agent
Claude Code
Model
Claude Sonnet 4.6

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.