agent-coordinator-swarm-init_skill

Overview

This skill initializes and configures multi-agent swarms for reliable, memory-coordinated workflows. It selects and optimizes swarm topology, allocates resources, and establishes inter-agent communication with mandatory shared-memory protocols. It enforces memory write/read conventions so every agent reports status, progress, artifacts, and completion to the coordination namespace.

How this skill works

On invocation the skill runs pre-hooks to log startup state and check for existing swarms, then chooses a topology (hierarchical, mesh, star, ring) based on task characteristics. It allocates compute and agent limits, creates shared memory channels in the "coordination" namespace, and configures event-driven message passing. Post-hooks write completion status and topology details, and the skill verifies that all agents are writing required memory keys during lifecycle events.

When to use it

• Bootstrapping a new multi-agent swarm for a coordinated project
• Setting up topology and resource limits before spawning specialized agents
• Enforcing consistent memory-based coordination across agents
• Preparing a swarm for distributed workloads like code analysis or feature development
• Reconfiguring topology and resources to optimize performance after monitoring

Best practices

• Choose topology to match workflow: hierarchical for structured control, mesh for peer collaboration, ring for sequential tasks
• Limit agents reasonably (typically 3–10) to avoid resource contention
• Require all agents to write initial status, step progress, shared artifacts, and completion to the coordination namespace
• Enable monitoring and performance analysis before production rollout
• Provide fallback topology/resource configs to handle initialization failures gracefully

Example use cases

• Initialize a mesh swarm for distributed code analysis that auto-optimizes topology
• Set up a hierarchical swarm of 8 agents for complex feature development with centralized coordination
• Prepare a star topology for centralized orchestration of API endpoint generation
• Enforce memory coordination for a spawned set of agents handing artifact sharing and dependency checks
• Reconfigure agent limits and network settings after performance monitoring recommends scaling