The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests content from documentation files that may contain instructions.
Ingestion points: The agent reads architecture documents like docs/ARCHITECTURE.md to perform updates.
Boundary markers: No delimiters or instructions are present to prevent the agent from obeying instructions embedded in the documents.
Capability inventory: The skill is designed to read and write documentation files.
Sanitization: No content validation or sanitization is performed on input documentation content.
[SAFE]: No hardcoded credentials, sensitive file access (e.g., SSH keys), or suspicious network operations were identified.
[SAFE]: No external dependencies are downloaded, and no remote code execution patterns are present.

docs-architecture