The Agent Skills Directory

[SAFE]: The skill provides instructions for the agent to use local documentation as a knowledge base. The actions described, such as reading and searching local text files within the project's documentation directories, are benign and aligned with the intended functionality.
[INDIRECT_PROMPT_INJECTION]: The skill processes content from local markdown and code snippet files which could theoretically contain instructions. However, as a documentation retrieval tool, this is its core function.
Ingestion points: SKILL.md identifies docs/content/ and docs/snippets/ as the primary data sources for the agent to read.
Boundary markers: None identified within the skill instructions.
Capability inventory: File reading (Read), searching (Grep, Glob), and fetching external API documentation (WebFetch) tools are authorized.
Sanitization: No specific sanitization or filtering logic is described for the ingested text.
[EXTERNAL_DOWNLOADS]: The skill references external documentation sources including ref.rerun.io and docs.rs. Both are recognized as trusted or well-known services for the vendor and the developer community respectively.

rerun-docs