The Agent Skills Directory

[SAFE]: The skill defines a structured operational workflow for project management, focusing on reading and updating local markdown-based implementation plans.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it derives its execution instructions from the content of project plan files.
Ingestion points: Project plan files located at plans/*.plan.md.
Boundary markers: The instructions do not define delimiters or specific markers to isolate the plan's data from the agent's core instructions.
Capability inventory: The skill utilizes Edit, Write, Agent, Task, and restricted Bash (git, ls) tools to execute the tasks described in the plans.
Sanitization: No explicit validation or sanitization logic is provided to verify the safety of instructions contained within the plan files.

plan-execution