DeepSeek Harness Plugins
Plugins that extend DeepSeek Harness (DSH) with vision, workflows, memory, web tools and more.
harness-pet plugin adds a pixel whale pet inside DeepSeek Harness, reacting to session signals with nine visual states, supporting drag, multi-language, and desktop window mode for fun interaction.
Provides attention reminders for the DeepSeek Harness Web UI, lighting up a badge, tab title count, and recolored favicon when a session needs user attention.
An independent DeepSeek Harness channel plugin that connects a WeCom AI Bot to persistent Harness agents via the official WebSocket long-connection SDK.
A self-control guard that intercepts high-confidence host-kill commands from the agent, teaches controlled exit tools, and logs an audit trail to prevent accidental termination.
dsh-orchestrator is a plugin for DeepSeek Harness that enables task decomposition and dispatch via a main agent, direct messaging between any agents, and per-worker model and thinking intensity configuration.
dsh-plugin-market adds a plugin marketplace to DeepSeek Harness, enabling users to browse, inspect evidence, and safely install or uninstall verified plugins within the Web UI.
Adds visual perception to text-only models in DeepSeek Harness via a free or custom vision API, generating and persisting descriptions without replacing the reasoning model.
dsh-stream-rules injects steering rules into agent workflows upon pattern matches on tool calls, controlling behavior without wasting context.
This plugin allows deleting sessions from a running DeepSeek Harness Web interface, including files, live store, and workspace records, without restarting the service.
dsh-calculator is a DeepSeek Harness plugin that displays API spend and account balance in a floating card on the Web GUI, supporting peak/off-peak pricing and real-time updates.
dsh-agent-message is a plugin for DeepSeek Harness that enables cross-session communication between different Agent sessions within the same process, supporting listing peers, sending messages, and checking delivery status.
ds-vision-plugin adds image input support to text-only DeepSeek models by automatically converting pasted images to text via a race among four vision models.