The Agent Skills Directory

[SAFE]: The skill implements a sophisticated multi-agent orchestration architecture using a coordinator and specialized worker roles (planner, executor). It utilizes session-scoped directories and a structured message bus to isolate task execution and maintain state.
[PROMPT_INJECTION]: The skill architecture is designed to ingest and process external data, which is standard for development orchestration tools.
Ingestion points: The system processes data from issue IDs, raw requirement text, and plan files. The executor role specifically reads solution artifacts generated by the planner role from the session's artifacts directory.
Boundary markers: Input data and planned solutions are interpolated into prompts for delegated tools (e.g., maestro delegate) using structured templates.
Capability inventory: The skill is granted access to the Bash, Write, Edit, and Agent tools, which are used for codebase modifications, git operations, and worker spawning.
Sanitization: The skill manages data flow through the multi-step pipeline, relying on the division of roles and the structured hand-off of solution artifacts between the planner and executor.
[COMMAND_EXECUTION]: The skill uses the Bash tool to perform standard development environment operations such as git commits and interacting with local workspace CLI tools. These operations are aligned with the skill's stated purpose of automating issue-based development.

team-planex