The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates extensive interaction with the local system via the bee CLI to manage project resources, including authentication, issue management, and direct API requests.
[EXTERNAL_DOWNLOADS]: Fetches additional command references and formatting instructions from the vendor's official GitHub and documentation sites (nulab.github.io and raw.githubusercontent.com/nulab).
[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources (Backlog issues, wiki pages, and pull request bodies) which could contain malicious instructions designed to influence the agent.
Ingestion points: Commands like bee issue view, bee pr view, and bee wiki view bring external content into the agent's context.
Boundary markers: None explicitly defined in the execution logic, though the text provides a manual security instruction to the agent.
Capability inventory: File system access (attachments), CLI execution, and network operations through bee api.
Sanitization: No automatic sanitization of the ingested content is described; the skill relies on the agent's internal safety filters and a specific text warning in the SKILL.md file.

using-bee