The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it retrieves and processes untrusted data from external project trackers.
Ingestion points: Ticket titles, descriptions, and comments are ingested from GitHub and Linear via MCP tools and the gh CLI.
Boundary markers: The instructions lack explicit delimiters or "ignore embedded instructions" warnings for the agent when processing external text.
Capability inventory: The skill's capabilities are limited to read-only tracker access and command execution through gh; write operations and persistence are explicitly forbidden.
Sanitization: The skill mandates that data be normalized into a "ticket brief," which provides basic structural filtering, though no formal sanitization of the input text is performed.
[COMMAND_EXECUTION]: The skill utilizes the gh (GitHub CLI) tool to retrieve information about issues and pull requests. These operations are limited to read-only "view" commands, such as gh issue view and gh pr view.

trackers