The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection vulnerability surface as it processes external data from an issue tracker to guide its planning and execution logic.
Ingestion points: Ticket titles, question bodies, and resolution comments are read from a repo issue tracker as defined in the 'The Map' and 'Tickets' sections of SKILL.md.
Boundary markers: The instructions lack explicit delimiting markers or system directives to ignore instructions that may be embedded within the ticket data.
Capability inventory: The agent can assign tickets, create issues, close issues, and launch subagents for research and prototyping.
Sanitization: There is no documented mechanism to validate, escape, or sanitize the content fetched from the tracker before it is processed by the LLM.
[COMMAND_EXECUTION]: The skill instructs the agent or user to execute /setup-matt-pocock-skills to initialize the tracking environment if one is not present. This introduces a dependency on external capability logic whose implementation and security parameters are not defined within this skill's scope.

wayfinder