The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches and executes markdown-link-check and markdownlint-cli from the official npm registry using npx --yes to perform validation tasks.
[COMMAND_EXECUTION]: The skill utilizes common shell utilities including find, grep, awk, sed, and wc to analyze file structures and document properties. It also suggests a pattern for automated validation that involves piping extracted code blocks into a shell (sh), which presents a risk if documentation contains untrusted content.
[DYNAMIC_EXECUTION]: The skill uses npx to dynamically download and execute code that is not part of the skill's static distribution.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data from project documentation, which could contain malicious instructions targeting the agent's behavior.
Ingestion points: Markdown files (.md), README files, and other documentation formats found in the project root or docs/ directory.
Boundary markers: The skill instructions do not specify the use of delimiters or 'ignore' instructions when reading content from external documentation files.
Capability inventory: The agent has access to Bash, Read, Edit, and MultiEdit tools, allowing it to execute scripts and modify files based on its analysis.
Sanitization: No explicit sanitization or validation of the content extracted from documentation files is mentioned before it is processed by the agent or used in command-line arguments.

documentation-expert