The Agent Skills Directory

Command Execution: The skill invokes system tools like git, rg, and make to scan the repository and verify documentation builds. These operations are standard for documentation synchronization and occur within the local environment.
Indirect Prompt Injection Surface: The skill processes source code and docstrings to identify documentation needs, which represents a surface where embedded instructions could potentially influence the agent's output. This risk is managed through a required report and user approval step before any file edits occur. 1. Ingestion points: Source files in src/agents/, examples/, and existing documentation. 2. Boundary markers: No specific delimiters are used to isolate untrusted content during reading. 3. Capability inventory: Writing to documentation files and mkdocs.yml, and executing make build-docs. 4. Sanitization: A mandatory human-in-the-loop approval workflow is used.

docs-sync