The Agent Skills Directory

[SAFE]: No security issues detected. The skill follow best practices for task decomposition and planning. It defines clear boundaries between planning, estimation, and execution, routing the user to appropriate specialized tools when necessary. The instructions are transparent and focus purely on the stated project management purpose.
[PROMPT_INJECTION]: The skill processes external project data such as issue lists and PRDs. While this is a known surface for indirect prompt injection, the risk is considered safe as it is intrinsic to the primary purpose of a planning assistant and the instructions emphasize structured output (YAML/Markdown) which helps mitigate accidental instruction obedience from data sources.
Ingestion points: SKILL.md (Step 2) identifies sources like repo/issues, prd/spec, gdd/playtest, and launch-notes.
Boundary markers: The skill instructs the agent to use specific markdown and YAML structures for the planning packets.
Capability inventory: The skill utilizes Bash, Read, Write, Edit, Glob, and Grep tools to manage project documentation.
Sanitization: No explicit input sanitization is mentioned, which is typical for LLM-based planning workflows.

task-planning