The Spawner skill enables the delegation of complex, long-running tasks to independent sub-agents using efficient model routing.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install dsb-spawner
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install dsb-spawner using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Spawner?
Spawner is a specialized module within the Openclaw Skills ecosystem designed to manage heavy workloads that exceed the scope of inline execution. By offloading tasks like significant file edits, full builds, or deep research to sub-agents, developers can maintain a clean primary context while ensuring specialized tasks are handled by dedicated roles using cost-effective model routing.
This skill is built for technical efficiency, prioritizing free model sequences such as Groq-Llama and Cerebras-Llama to maximize performance without draining premium token quotas. It provides a structured framework for defining sub-agent roles, exit criteria, and context injection, making it a cornerstone for complex multi-agent workflows.
Spawner Use Cases
- Performing large-scale code refactoring or multi-file edits that require more than two minutes of processing.
- Executing complex software build processes or running extensive test suites in an isolated environment.
- Conducting deep research or web browsing tasks through specialized browser-automator or researcher roles.
- Offloading technical writing and documentation generation to keep the main agent focused on logic.
How Spawner Works
- Role Definition: The user selects a specific persona for the sub-agent, such as a coder, writer, or reviewer.
- Context Selection: Only essential files and a task description under 2,000 tokens are injected into the sub-agent context.
- Model Routing: The system automatically routes the task through a hierarchy of free models (Groq-Llama → Cerebras-Llama → NIM-Llama → OpenRouter Scout).
- Execution and Monitoring: The sub-agent executes the task independently; the primary agent avoids polling and waits for an automatic announcement.
- Verification: Upon completion, the primary agent verifies the existence of the expected artifact and integrates only the output files, keeping transcripts out of the main context.
Spawner Setup
To integrate Spawner into your Openclaw Skills library, ensure your environment is configured for sub-agent management:
# Install the Openclaw CLI if not already present
npm install -g @openclaw/cli
# Enable the Spawner skill
openclaw skill add spawner
# Verify model provider connectivity for routing
openclaw check providers --include groq cerebras nim
Spawner Data Schema & Taxonomy
The Spawner skill utilizes a structured task template to ensure consistency across sub-agent deployments:
| Attribute | Description | Requirement |
|---|---|---|
| Role | One of: coder, writer, browser-automator, researcher, reviewer | Required |
| Context Files | List of file paths to be provided to the sub-agent | Minimum Required |
| Exit Criteria | Specific artifact (file path, URL, or test result) indicating success | Required |
| Token Limit | Maximum task description length | < 2,000 Tokens |
| Model Stack | Preferred routing: Groq → Cerebras → NIM → Scout | Automatic |
Spawner Advanced Features
- Intelligent Model Fallback: Automatically shifts to the next available model in the hierarchy if a 429 rate limit error occurs.
- Isolated Context Management: Prevents context bloat by ensuring sub-agent transcripts are never loaded into the primary agent's memory.
- Deterministic Exit Criteria: Forces a hard stop and error report if the sub-agent fails to produce the required artifact after two attempts.
- Concurrency Control: Native enforcement of a one-sub-agent-at-a-time rule to prevent resource contention and maintain execution clarity within Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
Active Self-Improvement
An autonomous self-optimization loop that enables AI agents to learn from past errors and successes to update their own internal protocols.
Task Extractor
A robust workflow skill designed to parse complex user messages into discrete, trackable, and verifiable action items.
UniProfit Mail Assistant
A specialized AI agent skill for sending verified emails through the UniProfit API using Openclaw Skills.
Tomato Fiction AI Short Drama Batch Generator
A comprehensive AI automation tool that converts multi-chapter novels into serialized short-form videos with automatic episode splitting and asset generation.
Podcast Digest
An advanced AI tool that converts podcast audio and transcripts into structured knowledge assets, interactive visual reports, and cross-episode matrices.
Rate Limit Guard
A protective layer for AI agents that prevents 429 rate limit loops and optimizes token usage during API failures.
Featured*
