The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted documentation files which could contain hidden instructions.
Ingestion points: SKILL.md instructs the agent to find and read all .md and .rst files in the repository.
Boundary markers: The instructions lack explicit delimiters or warnings to treat documentation content purely as data.
Capability inventory: The skill utilizes shell capabilities (find, grep, make).
Sanitization: No sanitization or validation of the documentation content is performed before processing.
[COMMAND_EXECUTION]: The skill suggests executing local build tools to verify documentation accuracy.
Evidence: references/validation-rules.md provides an example check using make help. Executing make triggers logic defined in the local Makefile, which could be used for malicious purposes if the repository is untrusted.

doc-validator