The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes codebase content to generate walkthroughs, which creates a surface for indirect prompt injection if the analyzed code contains malicious instructions.
Ingestion points: The skill instructions (Workflow Step 1) direct the agent to explore the repository, including READMEs, entry points, folder structures, and configuration files.
Boundary markers: There are no instructions provided to the agent to use delimiters or to ignore potential instructions embedded within the analyzed code or comments.
Capability inventory: The skill is limited to file system read operations for discovery and file system write operations restricted to the .tours/ directory for artifact creation.
Sanitization: The skill does not specify any sanitization, escaping, or filtering of the content read from the codebase before it is incorporated into the .tour file descriptions.

code-tour