The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by integrating untrusted external data into the orchestrator's decision-making process.
Ingestion points: Mission prompts in the plan_project and create_mission tools, and structured data returned by the get_report tool (SKILL.md).
Boundary markers: The instructions do not prescribe the use of boundary markers or instructions to isolate the data processed by the tools from the agent's core logic.
Capability inventory: The skill enables the agent to dispatch missions that execute code and perform git operations in isolated worktrees (dispatch_mission) (SKILL.md).
Sanitization: There is no mention of sanitization or validation for the user-supplied prompts or the contents of mission reports.

claude-devfleet