The Agent Skills Directory

PROMPT_INJECTION (LOW): Indirect Prompt Injection vulnerability surface detected. The skill's primary function is to read and analyze source code to generate documentation, creating a path for untrusted data to enter the agent's context.
Ingestion points: Phase 1 (Feature Analysis) and Phase 3 (Verification) read arbitrary files in the workspace using Read, Grep, and Glob tools.
Boundary markers: The prompt lacks instructions to distinguish between code-as-data and code-as-instructions, nor does it provide delimiters to encapsulate ingested file content.
Capability inventory: The skill is granted Bash, Write, and Edit permissions, providing an exploitation path if the agent follows instructions embedded within the analyzed files.
Sanitization: There is no logic provided to sanitize or filter natural language instructions found within code comments or string literals during the ingestion phases.

feature-docs