Analyzes Claude Code usage patterns and provides comprehensive recommendations. Runs usage analysis, discovers GitHub community resources, suggests CLAUDE.md improvements, and fetches latest docs on-demand.
SkillScore86Maintained
Automatic
◆AI Engineer pack
Last verified 8/6/2026
What it changes
Passed 16 of 19 test cases with the skill, and 7 of 19 without it.
Without skill36%
With skill86%
Checks it passed that the model alone missed
Recommends adding Bash to auto-allowed tools.
Specifies .claude/CLAUDE.md as the target file path.
Specifies that skill directories belong under .claude/skills/.
and 10 more in the full report
11 fixed8 passed either way3 still failing0 broke
Pass-rate lift
+50 pp
Tokens
+86%
Agent turns
0%
Response time
-59%
Model gemini-3.6-flashMeasured 8/6/2026v1 · latest is v219 of 22 cases passed with skill · 19 comparable
Tokens, turns and time describe resource use, not a dollar price.
This benchmark measured an earlier version of the skill.
See it in action
One of 11 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 11
Prompt
Before updating configuration options for tool execution approvals in Claude Code, which specific Anthropic documentation URL should be fetched to verify current settings options?
What was checked, without and with the skill
Checked
Without skill
With skill
References https://docs.claude.com/en/docs/claude-code/settings as the documentation URL to fetch.
Missed
Passed
case-07 · gemini-3.6-flash · 8/6/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md33 linessingle file
Claude Code Analyzer
Review Claude Code configuration as a tool-specific system. Current product behavior may change, so verify authoritative documentation when a recommendation depends on present syntax or availability.
Inventory
Inspect only configuration files and directories that are in scope. Typical areas include:
project and user CLAUDE.md instructions;
commands, agents, skills, hooks, and settings;
MCP server declarations and required credentials;
duplicated or conflicting instructions across scopes;
recurring manual workflows that may justify automation.
Do not assume helper scripts are installed. Use available read-only inspection tools and report inaccessible scopes.
Review Criteria
Scope: project-specific rules live with the project; personal preferences do not leak into shared configuration.
Clarity: instructions are concrete, non-conflicting, and tied to observable outcomes.
Context cost: large documents use progressive disclosure rather than loading unrelated guidance every turn.
Portability: paths, commands, dependencies, and environment requirements are explicit.
Safety: hooks and commands do not silently perform destructive or external actions.
Maintainability: duplicated rules have one source of truth and obsolete entries are removed.
Evidence: recommendations refer to actual files, behavior, or documented product capabilities.
Migration or Comparison
When translating configuration to Codex or another agent, preserve intent rather than copying filenames mechanically. Map project instructions, reusable skills, tool integrations, invocation rules, and permission boundaries to the destination's supported concepts. Flag features without an equivalent.
Deliverable
Report findings by impact, with file paths and concrete remediation. Separate confirmed problems from optional improvements and product-version uncertainty. Do not edit configuration unless the user requests changes.