Spawn specialized subagents with scoped file ownership and permission checks.
npx skills add https://github.com/c-daly/agent-swarm --skill spawn-c-daly
Please help me install this Agent Skill. Skill: spawn Source: https://github.com/c-daly/agent-swarm/tree/main/skills/spawn Command: npx skills add https://github.com/c-daly/agent-swarm --skill spawn-c-daly
This Skill streamlines the creation and management of subagents to handle specialized tasks, enabling efficient workload distribution and better context isolation.
Use the spawn skill to delegate a requirements-gathering task to an Explorer subagent and a Researcher to gather docs.
High-intent search queries and answers about installing and using this skill.
Delegating tasks to subagents automates workload distribution by spawning Explorer, Researcher, Architect, Implementer, and Reviewer agents for specific responsibilities. Define clear task prompts, assign scoped file ownership, specify output formats, and enforce permission checks to maintain safety and traceability across the orchestrated delegation.
Orchestration spawns specialized subagents as needed rather than overloading a single agent with all work. Each subagent handles a defined scope—code discovery, research, architecture, implementation, or review—with isolated context and independent reporting, reducing errors and improving task clarity.
Task delegation with subagents fits projects requiring parallel work streams, external research, code discovery, and specialized expertise. When a task requires both web research and codebase analysis simultaneously, spawning dedicated agents prevents context collision and accelerates completion.
Configure per-task file ownership by defining which files each subagent can access and modify, then enforce permission checks at spawn time. This maintains safety in multi-agent contexts by preventing unintended file modifications and ensuring traceability across delegated work.
Specify structured output formats for each subagent task before spawning to enable automated verification and handoff between agents. Clear formats allow orchestrators to validate results, extract data reliably, and route outputs to downstream agents without manual parsing.
Yes; task delegation isolates specialized work in scoped subagent contexts, containing risk from experimental code, external data, or complex analysis. Guardrails and output verification ensure failed or unexpected subagent work doesn't corrupt the main workflow.
Shockwave2kOrchestrate Claude-led agent teams with spawn rules and cleanup protocols.
SeventeenthEarthDelegate codebase searches to GLM-4.7 via a background subagent.
kangarookingCoordinate multi-agent delegation with lifecycle gates and context isolation.
midsphere-aiSpawn parallel sub-agents with configurable depth and memory isolation.
