Interface overview
cwbcheng/dsh-knowledge-graph · model-adapters
DSH Cordis plugin: Convert any source text into an AI knowledge graph (facts/inferences/concepts/definitions/examples/counterexamples/rules) with bidirectional links between the graph and the original text.
★ 2GitHub stars
0Forks
2026-08-16Last updated
JavaScriptLanguage
MITLicense
Key features
- DSH (DeepSeek Harness) Cordis plug-in: Use AI to disassemble any piece of data text - or an AI session execution track - into a knowledge graph, and bidirectionally position between the knowledge graph and the original text.
- AI asynchronous splittingEnter any text (chapter, technical document, study notes...), call LLM in background task mode, and return a knowledge graph in about 15-40 seconds.
- Dynamic installation (method A/B): Repeat method A after the warehouse is updated - let the Agent re-read the two source files and cordis_define (add a new Package under the same plug-in), and then cordis_run (update mode) to switch to the new version
- Dynamic installationOpen the Cordis Plugin panel → click Stop in the plug-in row to suspend use
- Long documents are automatically processed in batches, and polling results are automatically restored after refreshing/reopening the window.
Requirements
- DSH Web (dsh web) has been started and entered into any session
- The AI model provider is configured in the environment (Settings → Models, or agentDefaultModel).
Install command
dsh plugin --profile web add github:cwbcheng/dsh-knowledge-graph