Vibe-Mathematics
Run the following command in DeepSeek Harness:
dsh plugin install ChongCyrus/Vibe-Mathematics
Paste the following prompt into your AI chat to install this plugin:
In DeepSeek Harness, install the plugin package from https://github.com/ChongCyrus/Vibe-Mathematics.
About this plugin
Vibe-Mathematics addresses a difficult but familiar problem in mathematical reasoning: long proof chains can easily drift, miss assumptions, or produce answers that look complete but cannot be trusted. The plugin turns a mathematical task into a collaborative workflow where agents can explore different directions, solve step by step, verify conclusions independently, and save reusable findings along the way.
At its core, it provides a set of multi-agent presets for DeepSeek Harness, covering a classic pipeline, a probability-driven JSON data layer, a Markdown knowledge-base workflow with planning agents, and an experimental self-organizing resident-agent mode. Across these presets, users can resume interrupted runs, intervene manually, review progress, and control the workflow through natural language instead of micromanaging each step.
It is best suited for mathematical researchers, educators, AI engineers, and users experimenting with multi-agent reasoning. If you need traceable intermediate claims, reusable methods, paper-like knowledge accumulation, or independent verification before accepting a conclusion, Vibe-Mathematics offers a more structured and controllable way to turn a single model answer into an auditable problem-solving process.
Screenshots
Use Cases
- Break complex math problems into multiple solving directions
- Cross-validate propositions and solutions with multiple agents
- Save proofs, methods, and progress as reusable knowledge
Best For
- Mathematical researchers and educators
- DeepSeek Harness multi-agent workflow users
- Engineers and experimenters needing auditable reasoning
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.