dsh-kimicode-swarm
Run the following command in DeepSeek Harness:
dsh plugin install hongyue0721/dsh-kimicode-swarm
Paste the following prompt into your AI chat to install this plugin:
To install this plugin in DeepSeek Harness, use the command dsh plugin install hongyue0721/dsh-kimicode-swarm, with the full source code available at https://github.com/hongyue0721/dsh-kimicode-swarm.
About this plugin
In DeepSeek Harness workflows, when tasks are split into multiple independent subtasks, traditional serial execution is inefficient and hard to scale, leading to prolonged processing times and underutilization of parallel computing resources.
The dsh-kimicode-swarm plugin solves this by integrating Kimi Code Swarm's master-slave multi-agent parallel architecture. It offers the swarm_batch tool, enabling users to batch-dispatch up to 128 subtasks at once for parallel execution by sub-agents, with adaptive scheduling mechanisms like exponential backoff and concurrency capacity adjustments to boost end-to-end efficiency by up to 4.5 times. Real-time progress bars in the chat interface visually display each sub-agent's status (queued, running, completed), making monitoring straightforward.
This plugin is ideal for users handling numerous independent tasks, such as developers performing batch code generation, test execution, or data analysis. It simplifies parallel task management, supports flexible model allocation based on task difficulty, and enhances efficiency and user experience for exploring new boundaries or automating repetitive workflows.
Screenshots
Use Cases
- Batch processing independent subtasks to boost efficiency
- Automating test execution and code generation
- Accelerating data analysis or exploratory tasks
Best For
- Developers requiring parallel task processing
- Engineers handling batch code or test work
- Data scientists managing repetitive workflows
Related Plugins
A method pack that makes AI coding agents plan against your real baseline, prove completion with fresh evidence, and reduce reworks and unsafe changes.
Turns the DeepSeek Harness session into a captain that builds a durable sub-agent team, splits goals into dependency-aware tasks, and coordinates work via direct messages and a live Web UI.
Gives coding agents design judgment, letting Claude Code, Cursor, and 70+ agents generate and iterate high-quality UI, presentations, and graphics right from the terminal.
Run the Pi ecosystem's plugins on DeepSeek Harness, unmodified, via a compatibility layer that implements Pi's public extension ABI on DSH's native services.