Coordinate multiple AI agents on shared subtasks using SQLite atomic claiming.
npx skills add https://github.com/ssccio/cc-on-web-base --skill swarm-ssccio
Please help me install this Agent Skill. Skill: swarm Source: https://github.com/ssccio/cc-on-web-base/tree/main/skills/swarm Command: npx skills add https://github.com/ssccio/cc-on-web-base --skill swarm-ssccio
It eliminates slow, error-prone single-agent work by coordinating multiple agents to process many subtasks in parallel while safely dividing ownership of work items.
Use Case Examples:
Provide the command /swarm 5:executor "fix all TypeScript errors" to spawn five coordinated executors that claim and work through the decomposed file-level subtasks.
High-intent search queries and answers about installing and using this skill.
Multi-agent coordination for parallel codebase automation uses SQLite-based atomic task claiming to let multiple agents draw from a shared subtask pool without race conditions. Agents claim, process, and complete decomposed file-level tasks concurrently with transactional state transitions.
Fault tolerance in parallel agent execution uses lease-based ownership with heartbeat monitoring. If an agent stalls, stale-claim cleanup automatically recovers the task, ensuring batch security or design reviews complete reliably without manual intervention.
Parallel task claiming enforces an orchestration limit of 1 to 5 agents. This constraint ensures safe multi-agent coordination while processing decomposed subtasks like fixing TypeScript type errors across multiple files concurrently.
Start a parallel documentation sprint by running a command like `/swarm 5:executor` with your task description. This spawns five coordinated agents that automatically decompose the work, claim file-level subtasks, and track pending, claimed, done, and failed states.
Yes, concurrent multi-agent execution provides progress visibility by tracking pending, claimed, done, and failed task counts. It monitors which agent owns which task and produces a clear final summary upon deterministic completion signaling.
The metadata does not specify external SQLite setup requirements. SQLite-based atomic claiming and transactional state transitions are integrated directly into the multi-agent coordination process to manage concurrent subtask ownership safely.
suhwanOrchestrate coordinated agents on a shared SQLite-backed task list.
leocostaDetect unverified completion claims by comparing claimed results to build, test, and typecheck evidence.
bajajvinamrOrchestrate multi-agent swarms across distributed networks with fault-tolerant coordination.
Veloxia-agencyCoordinate parallel subagents in isolated git worktrees for shared tasks.
