New agent startup checklist for Agent Mail and Beads. Use when starting a new agent session, when the user says "prime" or "startup", or when beginning work on a multi-agent project.
The "prime" skill provides a complete startup checklist for initializing a new agent session within Agent Mail and Beads environments. It is designed to standardize and automate the setup process whenever a new session begins, when a user explicitly invokes startup commands such as "prime" or "initialize," or when coordinating work across multiple agents. By consolidating identity registration, environment orientation, and coordination steps into a structured flow, this skill ensures that every session begins in a consistent, traceable, and collaboration-ready state.
The skill executes directly without spawning subagents, as each step consists of lightweight command sequences rather than analytical workloads. Its core capabilities include registering an agent identity using either a preferred macro-based method or manual tool calls, configuring contact policies for coordination, updating the terminal context, reviewing project documentation files, inspecting recent session history, retrieving contextual patterns, checking repository status, discovering active agents, and processing inbox messages. It also includes structured message categorization (urgent, informational, acknowledgement-required) and coordination safeguards such as respecting file reservations from other agents.
This skill is intended for agents operating in structured multi-agent projects that rely on Agent Mail (MCP) and related tooling. It is particularly useful for collaborative development environments where consistent initialization, communication visibility, and repository awareness are essential. Teams working on coordinated tasks, shared repositories, or session-based workflows benefit from its standardized startup procedure, which reduces ambiguity and prevents misalignment at the beginning of each work cycle.
It should be used at the start of a new agent session, when the user explicitly invokes "prime," "startup," or "initialize," or when beginning coordinated work in a multi-agent project.
No. The skill executes directly because each step consists of simple command sequences, and subagent overhead would exceed the actual work required.
It uses Agent Mail (MCP) tools such as session macros, agent registration, contact policy configuration, inbox retrieval, and messaging, along with Bash commands for context retrieval, session history search, task discovery, and git repository inspection.
Yes. The macro-based startup method is preferred because it consolidates identity registration and initialization into a single call.
The skill requires setting the contact policy to open for coordination, checking inbox messages (including acknowledgement requirements), identifying active agents, and respecting file reservations before beginning work.
Quick Setup:
.claude/skills/Repository
Mburdo/knowledge_and_vibes