dsh-bridge
Run the following command in DeepSeek Harness:
dsh plugin install baixianger/dsh-bridge
Paste the following prompt into your AI chat to install this plugin:
Install this plugin in DeepSeek Harness by running dsh plugin install baixianger/dsh-bridge; full source code is available at https://github.com/baixianger/dsh-bridge
About this plugin
Running multiple agent sessions on a single DeepSeek Harness host often leaves them as information silos—unable to discover one another or hand off context. dsh-bridge closes that local communication gap by providing a session discovery and messaging layer inside a single host process, so every node in your workflow can actually talk to the others.
Three tools cover the essentials: session_list surfaces every active session with its current state; session_send delivers a message by exact ID or human-readable title, gracefully handling idle wake-ups, queued work for running agents, and merged cold-resume operations; session_messages returns a bounded slice of the recent delivery log. The delivery engine respects the full session lifecycle—archived sessions reject messages, in-flight cancellations prevent stale follow-ups, and every path shares a unified acknowledgement and audit trail.
Built for developers orchestrating multi-agent pipelines on one host, and for plugin authors who need a local transport foundation beneath DSH Chat rooms or DSH Weave cross-host delivery. It runs entirely in memory, requires no external service, and is ready the moment the plugin loads.
Screenshots
Use Cases
- Orchestrating multiple agent sessions on one host to collaborate on complex tasks
- Building a local session discovery and delivery foundation for new plugins
- Monitoring and auditing inter-agent message records and delivery status
Best For
- Developers orchestrating multi-agent workflows on a single host
- Plugin authors building a local transport layer for DSH Chat or Weave
- Operators who need to debug and audit inter-agent message flows in production
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.