The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill heavily relies on data ingested from an external issue tracker (titles, bodies, and decision gists) to determine its next actions and 'clear the fog' of planning. Malicious instructions placed within these issues could influence the agent's behavior.
Ingestion points: Loading the map issue, fetching child ticket bodies, and reading decision comments in the 'Decisions-so-far' section.
Boundary markers: Absent. The instructions do not specify the use of clear delimiters or instructions to ignore embedded prompts when reading issue content.
Capability inventory: The agent can create and wire issues, post resolution comments, create 'research' branches with code/assets, and spawn subagents.
Sanitization: Not identified. The agent processes issue text directly for decision-making.
[EXTERNAL_DOWNLOADS]: The skill instructs the user to run /setup-matt-pocock-skills if a tracker is not provided. This references an external setup utility from a third-party source not identified as a standard or trusted vendor.
[DYNAMIC_EXECUTION]: The skill involves spawning 'research subagents' to resolve tickets in parallel. While this is a feature of the orchestration, it represents the dynamic creation of new execution contexts based on the results of the planning phase.

wayfinder