convoy-management
Scannednpx machina-cli add skill a5c-ai/babysitter/convoy-management --openclawConvoy Management
Overview
Manage the full lifecycle of Gas Town convoys: creation from a goal, bead decomposition, agent assignment, progress tracking, and landing (merge). Convoys are the primary work-order unit wrapping related beads.
When to Use
- Starting a new multi-agent work effort
- Coordinating parallel bead execution across agents
- Tracking progress of distributed work
- Landing (merging) completed convoy results
Process
- Create convoy from goal or MEOW decomposition
- Decompose into beads (persistent) and wisps (ephemeral)
- Assign beads to Crew (long-lived) or Polecats (transient)
- Track progress via hooks and heartbeats
- Verify all beads complete with quality checks
- Land convoy by merging bead branches
Key Concepts
- Convoy: Primary work-order wrapping related beads
- Bead: Git-backed atomic work unit (issue/task)
- Wisp: Ephemeral bead destroyed after successful run
- GUPP: If there is work on your Hook, YOU MUST RUN IT
Agents Used
agents/mayor/- Creates and coordinates convoysagents/crew-lead/- Persistent collaborator on beadsagents/polecat/- Transient worker for individual beads
Tool Use
Invoke via babysitter process: methodologies/gastown/gastown-convoy
Source
git clone https://github.com/a5c-ai/babysitter/blob/main/plugins/babysitter/skills/babysit/process/methodologies/gastown/skills/convoy-management/SKILL.mdView on GitHub Overview
Convoy management orchestrates the full lifecycle of Gas Town convoys: creation from a goal or MEOW decomposition, bead decomposition into persistent beads and ephemeral wisps, and landing (merging) of bead branches. Convoys are the primary work-order unit that wraps related beads for coordinated multi-agent execution. This approach enables scalable, trackable orchestration of distributed tasks across crews and polecats.
How This Skill Works
Create a convoy from a goal or MEOW decomposition, then decompose into beads (persistent) and wisps (ephemeral). Assign beads to Crew (persistent) or Polecats (transient) and monitor progress with hooks and heartbeats, performing quality checks before landing by merging bead branches.
When to Use It
- Starting a new multi-agent work effort
- Coordinating parallel bead execution across agents
- Tracking progress of distributed work
- Landing (merging) completed convoy results
- Decomposing goals into convoys and beads
Quick Start
- Step 1: Create convoy from a goal or MEOW decomposition
- Step 2: Decompose into beads (persistent) and wisps (ephemeral)
- Step 3: Assign beads to Crew (persistent) or Polecats (transient), monitor with hooks/heartbeats, then land by merging bead branches
Best Practices
- Define clear convoy goals at creation
- Keep beads atomic and Git-backed (issues/tasks)
- Assign persistent beads to Crew and transient beads to Polecats
- Use hooks and heartbeats to monitor progress
- Perform upfront quality checks before landing
Example Use Cases
- A convoy is created to orchestrate a data-labeling sprint with multiple beads assigned to Crew-leads; progress is tracked via hooks and heartbeats, and the convoy lands after QC passes.
- Parallel API integration tasks are decomposed into beads and dispatched to Polecats for concurrent execution.
- A complex model-training task is decomposed into beads (persistent) and wisps (ephemeral), with progress tracked until completion.
- After all beads complete and pass quality checks, the convoy lands by merging bead branches to produce the final results.
- A Hook-specific GUPP is triggered to ensure no work remains on the hook before landing.
