The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches documentation files (llms.txt, llms-full.txt) from the vendor's official documentation domain (docs.bitdrift.io). This is the intended purpose of the skill and follows safe practices by using fixed origins.
[INDIRECT_PROMPT_INJECTION]: The skill processes external documentation content which could theoretically contain malicious instructions. However, the skill includes a 'Trust boundary' section that explicitly instructs the agent to treat fetched data as reference material only and forbids executing commands or changing secrets based on that content, effectively mitigating the surface risk.
[COMMAND_EXECUTION]: The skill uses curl to fetch documentation. The commands are static, target a specific trusted vendor domain, and do not incorporate unvalidated user input into the shell execution line.

bd-docs