The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes content from external source files (Python, JavaScript, TypeScript) using a provided extraction script, which constitutes an indirect prompt injection surface. Instructions embedded in code comments within these files could potentially influence the agent's behavior or output.
Ingestion points: The scripts/extract_comments.py script reads user-specified source files from the filesystem.
Boundary markers: The instructions do not define specific delimiters or guidelines to ignore instructions embedded within the extracted comments.
Capability inventory: The agent is instructed to execute the extraction script and analyze its output. It possesses file reading and command execution capabilities.
Sanitization: The skill does not implement sanitization, validation, or filtering of the extracted comment text before processing it.

code-comments