The Agent Skills Directory

[SAFE]: The skill is designed for local project analysis. It uses Grep to calculate code metrics (line counts and branching) and reads project documentation to inform its estimates. It does not possess capabilities for network exfiltration or modifying the filesystem.
[INDIRECT_PROMPT_INJECTION]: The skill accepts an external task-description as input. While this data is untrusted, the skill's restricted toolset (Read, Glob, Grep) and its read-only nature prevent any malicious instructions in the task description from performing harmful actions like remote code execution or data exfiltration.
[COMMAND_EXECUTION]: The skill uses the Grep tool to perform regex-based searches. The patterns provided in the instructions are standard for measuring code complexity and do not attempt to bypass tool constraints or execute shell commands.

estimate