The Agent Skills Directory

[Indirect Prompt Injection Surface]: In Phase 1, the skill fetches and processes content from user-provided documentation URLs to extract technical details for the mission specification. As with any agent that processes external web content, there is a potential for instructions embedded in those documents to influence the agent's behavior. The skill incorporates a mitigation by presenting the drafted issue body to the user for confirmation before it is posted.
[System Command Execution]: The skill utilizes the GitHub CLI (gh) for repository interactions, such as checking authentication status, listing milestones, and creating issues. These commands are restricted to the microsoft/agent-academy repository and are consistent with the skill's stated purpose of managing GitHub project content.
[Secure File Handling]: The skill drafts issue bodies to temporary files in /tmp/ rather than the active workspace and ensures they are deleted after the issue is created. This follows best practices for handling transient data and minimizes the footprint on the local filesystem.

create-mission-spec