Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →miosa-osa/spawn
Launch a specialized agent for a subtask. Assigns identity, workspace scope, available commands, and communication protocol. Supports tmux-based (visual) or subprocess (headless) execution backends. The primary way to parallelize work.
Passed 22 of 22 test cases with the skill, and 4 of 22 without it.
and 23 more in the full report
18 fixed4 passed either way0 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 18 test cases the skill fixed. Same model, same prompt, with and without the skill.
We need to run our 'researcher', 'analyst', and 'writer' library agents at the same time to create a benchmark report using tasks defined in 'report-plan.md'. Provide the exact CLI command line to spawn all three simultaneously.
| Checked | Without skill | With skill |
|---|---|---|
| The command lists researcher, analyst, and writer positionally. | Missed | Passed |
| The command includes the --parallel flag. | Missed | Passed |
The instructions your model reads when the skill loads.
Launch a specialized agent for a focused subtask.
Create and start an agent runtime dedicated to a specific subtask. Each spawned agent gets its own identity, scoped workspace access, command allowlist, and communication channel back to the orchestrator. Use this to parallelize work, isolate concerns, or assign specialized tasks to purpose-built agents.
# Spawn an agent from the library
/spawn researcher --task "Find pricing benchmarks for AI SaaS"
# Spawn with explicit backend
/spawn coder --backend tmux --task "Implement the search endpoint"
# Spawn with scoped workspace access
/spawn analyst --scope node:money-revenue --task "Generate Q1 revenue report"
# Spawn with budget limit
/spawn writer --budget 50000 --task "Draft the pitch deck content"
# Spawn with custom identity
/spawn --name "pricing-researcher" --role "Market Analyst" --task "Compare competitor pricing"
# Spawn multiple agents
/spawn researcher analyst writer --parallel --tasks-from plan.md
| Flag | Type | Default | Description |
|---|---|---|---|
<agent> | positional | required | Agent name from library, or custom with --name |
--task | string | required | The specific task to assign |
--name | string | auto | Custom agent name (overrides library name) |
--role | string | from library | Role description for the agent's identity |
--backend | enum | subprocess | subprocess (headless), tmux (visual), docker (isolated) |
--scope | string | full | Workspace scope: full, node:<name>, read-only, glob pattern |
test cases passed, vs 4 / 22 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.