The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses a standard surface for indirect prompt injection as it facilitates the analysis and refinement of user-provided documentation and templates. * Ingestion points: The agent is instructed to fetch and analyze existing documents or user-provided templates (SKILL.md, Part 2). * Boundary markers: There are no explicit separators or instructions to ignore embedded commands within user content are provided. * Capability inventory: The skill includes a markdown validation script and capabilities for document drafting/writing. * Sanitization: No specific filtering or sanitization of external content is mentioned.
[COMMAND_EXECUTION]: The skill includes a local utility script, scripts/doc-structure-validator.py, which is intended to be used by the agent to perform static structural analysis of markdown files. * Evidence: The script uses standard Python libraries (argparse, os, re, json, pathlib) to validate markdown files for issues like heading hierarchy and empty sections. It does not invoke external commands, subshells, or perform any network operations.

documentation-authoring