deepseek-ai/deepseek-harness

DeepSeek Harness: Everything is a Plugin.

View on GitHub ↗Jump to charts ↓

Summary Information

Updated 25 minutes ago
Added to GitGenius on August 13th, 2026
Created on August 13th, 2026
Open Issues & Pull Requests: 0 (+0)
GitHub issues: Disabled - open counts may still include pull requests.
Number of forks: 8,564
Total Stargazers: 93,445 (+6)
Total Subscribers: 358 (+0)

Repository Insights (GitGenius)

Most active contributors

Sign in to see contributor activity.

Related repositories by overlapping contributors

No overlapping-contributor repos identified yet.

Charts & Analytics

Fetching additional details & charts...

Issue Activity (beta)

GitHub issues are disabled for this repository, so issue analytics and the issue explorer are not available.

Detailed Description

DeepSeek Harness is an open-source agent harness that uses a plugin-based architecture powered by Cordis.

The tool addresses the need for a flexible, composable framework for building agents by implementing a design where every component functions as a plugin. This architecture, grounded in the Cordis framework and its spatiotemporal composability paradigm, allows developers to extend and customize agent behavior through discrete, pluggable modules rather than monolithic structures.

Developers should adopt this tool if they are building agents and want a highly modular, extensible foundation. The plugin-first design makes it suitable for projects where functionality needs to be added, removed, or swapped without modifying core systems. The project provides a Web UI served locally by default and includes comprehensive documentation covering usage, development, and agent-specific implementation patterns. Those evaluating adoption should be aware that the tool is in developer preview with rapid iteration and compatibility-breaking changes expected.

The project shows active engagement with its community through GitHub Discussions and maintains a Discord community channel. Development documentation is structured across multiple guides covering architecture, development setup, and agent implementation, indicating sustained attention to onboarding and knowledge transfer. The codebase is written in TypeScript and encourages plugin discoverability through a standardized topic tag for plugin repositories.