The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the GitHub CLI (gh) to retrieve issue and pull request data via shell commands like gh issue view and gh pr view. These operations are limited to data retrieval with JSON output formatting.
[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted content from external ticket descriptions and comments from Linear and GitHub.
Ingestion points: External data enters the agent context through Linear ticket bodies and GitHub Issue/Pull Request descriptions and comments.
Boundary markers: The skill mandates the use of a specific "Ticket Brief" template to isolate and structure the external data under clearly defined markdown headers (# Ticket, # Acceptance, # Constraints).
Capability inventory: The skill's capabilities are strictly limited to read-only tools and CLI commands. It explicitly forbids any write actions, such as save_issue, update_issue, close, or comment.
Sanitization: The skill instructs the agent to selectively pull specific fields (acceptance criteria, priority, labels) and to ignore "pure chatter" in comments, which reduces the surface area for malicious instructions hidden in external text.

trackers