Runs Claude Code non-interactively for code analysis and implementation tasks. Use when the user asks to delegate repository work to Claude Code or compare its result with another coding agent.
71
86%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
Passed
No findings from the security scan
command -v claude. If it is missing, stop and ask the user to install Claude Code in Code Mate.claude -p "<prompt>" --output-format jsonPass the prompt as one quoted argument. Parse the JSON result and report stderr and the exit status on failure. Never start the interactive REPL or /login flow.
If Claude reports missing login, API credentials, or provider configuration, stop and ask the user to finish Claude Code setup in Code Mate. Never request, read, print, or copy credentials.
Keep tool access read-only by default. Add only the smallest necessary --allowedTools entries when the task explicitly requires tools, and grant edit or shell tools only when the user explicitly requested workspace changes. Do not use a permission-bypass mode.
Example: for a review request, ask Claude to inspect the current repository without changing files, run the command above, then summarize the parsed JSON result.
113ef7e
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.