deepseek-harness: Agent framework with plugin-based architecture
Build agents where every part,model, tools, UI, storage,swaps without editing code. Choose capabilities at runtime or write your own plugins to compose exactly what you need.
Compiled automatically from public sources — the company's own posts, launch material and reporting, all linked at the bottom. The “what using it looks like” section is a walkthrough built from public documentation, not a first-hand review. If this is your company, claim the page and correct anything wrong.
Build agents where every part,model, tools, UI, storage,swaps without editing code. Choose capabilities at runtime or write your own plugins to compose exactly what you need.
What it does
Plugin-based framework for AI agents. All capabilities,models, tools, skills, sessions, sandboxes, storage, scheduling, UI,are provided as replaceable plugins coordinated by a Cordis kernel.
Who it's for
Developers building AI agents who need composition and flexibility without forking code.
Where it fits
- Swap model or tool mid-project
- Mix execution modes for different scenarios
- Write custom capabilities as plugins
The short read
- Plugin architecture replaces fixed agent components
- Change any capability in configuration, no code edit
- Runs in four modes: standard, code-gen, minimal, creative
The idea
If every agent capability becomes a plugin, configuration becomes the infrastructure layer and teams can compose exactly what they need without rewriting core.