What problem does it solve? Coordinating multiple AI agents in a session requires consistent communication rules and a reliable way to pass tasks and reports between agents without losing or overwriting messages. ## Core Features & Use Cases - Communication Standards: Defines how agents talk to users (visible reasoning, natural language, terminology boundaries) and how they write structured artifacts. - File-Based Message Bus: Specifies a protocol using per-agent directories under .apm/bus/ with task.md and report.md files, including write ownership, collision safety, and identity validation rules. - External Agent Integration: A companion guide lets non-APM agents join the bus by creating their own slug directory and exchanging task and report files. - Use Case: A Manager agent assigns work to a Frontend Agent by writing to .apm/bus/frontend-agent/task.md; the Worker reads it, completes the task, and reports back via report.md without message collisions. ## Quick Start Ask the agent to set up the APM message bus directories and explain how to send a task message to a worker agent.
