The Agent Skills Directory

[PROMPT_INJECTION]: The skill acts as an orchestration layer that processes natural language prompts to define 'missions' for sub-agents, creating a surface for indirect prompt injection if those prompts are derived from untrusted external data. \n
Ingestion points: The prompt parameters in plan_project, create_mission, and dispatch_mission tools within SKILL.md. \n
Boundary markers: No delimiters or instructions to ignore embedded commands are specified for the mission prompts. \n
Capability inventory: Spawns autonomous agents with 'full tooling' and access to git worktrees. \n
Sanitization: No input validation or content filtering is described in the tool definitions. \n- [COMMAND_EXECUTION]: The skill is designed to manage and dispatch agents that execute arbitrary code and commands within isolated git worktrees to perform coding tasks. \n- [NO_CODE]: The skill consists exclusively of markdown documentation and tool definitions in SKILL.md, with no accompanying executable scripts or binary files.

claude-devfleet