The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill defines a pipeline that ingests external codebase content, which constitutes a vulnerability surface for indirect prompt injection. Ingestion points: Exploration and writer agents read the full content of the project codebase (SKILL.md). Boundary markers: The instructions do not specify delimiters to isolate code content from agent instructions. Capability inventory: The agents possess the capability to read project files and write multiple markdown files to a .codebase-map/ directory (SKILL.md). Sanitization: The writing guidelines do not mention sanitization or filtering of codebase content.
[SAFE]: The skill consists of formatting rules, tone guidelines, and Mermaid diagram standards. It does not include executable code, network operations, or hardcoded credentials. All described activities are local to the project environment and consistent with the skill's stated purpose of documentation generation.

codebase-mapper