The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands to facilitate codebase exploration and document management. It defines helper functions _finder and _rg to abstract the use of fd/find and rg/grep for searching files. It also uses mkdir to create directories, wc to verify outputs, and git for optional commits of the generated documentation.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core function of reading and processing untrusted codebase content.
Ingestion points: Exploration commands like cat, rg, and fd are used to read file contents across the project repository.
Boundary markers: The prompts for the sub-agents do not specify delimiters or include explicit instructions to disregard potential commands embedded within the analyzed source files.
Capability inventory: The orchestration involves spawning background processes via the Task tool, writing persistent files with the Write tool, and executing git commands.
Sanitization: There is no evidence of sanitization or safety validation for content ingested from the codebase before it is processed by the agents.

codemap