The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill communicates with the vendor's official domain (unbrowse.ai) to fetch documentation data and execute tools. This is the primary function of the skill and utilizes the service's own infrastructure.
[INDIRECT_PROMPT_INJECTION]: The skill processes external content from docs.rs which could theoretically contain instructions targeting the agent. This is a standard risk for tools designed to ingest web content.
Ingestion points: Tools docs_rs__get_search, docs_rs__get_releases_search, and docs_rs__read_page ingest documentation content as described in SKILL.md.
Capability inventory: The skill provides generic task execution and cloud browsing capabilities through unbrowse.run and unbrowse.browse.* tools.
Boundary markers: No explicit delimiters are mentioned in the instructions to separate documentation data from instructions.
Sanitization: No specific sanitization or filtering logic is described for the content retrieved from external pages.
[COMMAND_EXECUTION]: The README.md and SKILL.md files provide standard setup commands for platform-specific skill installation (claude mcp add, codex mcp add, npx skills add).

docs-rs