The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources, including URLs, issue tracker comments, and user-provided specification files. This represents an attack surface where maliciously crafted external content could attempt to influence the agent's logic during the ticket breakdown process. However, this is a standard risk for tools that aggregate external documentation, and no specific exploitation patterns were found.
[DATA_EXFILTRATION]: The skill writes ticket data to a local .scratch/ directory and publishes issues to external platforms (GitHub, Linear). These operations are core to the skill's stated purpose and are performed based on the user's configuration and approval flow.
[COMMAND_EXECUTION]: The skill mentions external tools like /setup-matt-pocock-skills and /implement. These are referenced as external dependencies for configuration and task execution but do not represent arbitrary command execution within the skill itself.

to-tickets