Build vector-indexed knowledge bases from project files, database metadata, and validated SQL corpora.
npx skills add https://github.com/Datus-ai/Datus-agent --skill build-kb
Please help me install this Agent Skill. Skill: build-kb Source: https://github.com/Datus-ai/Datus-agent/tree/main/datus/resources/skills/build-kb Command: npx skills add https://github.com/Datus-ai/Datus-agent --skill build-kb
This Skill eliminates the manual, error-prone work of building and maintaining searchable, domain-aware knowledge bases for data projects, which is critical for ensuring LLM-generated SQL is accurate and aligned with business context.
Use the build-kb skill to generate a vector knowledge base for the sales domain, including semantic models for orders tables, metrics for monthly revenue, and all validated sales queries from the queries/ directory.
High-intent search queries and answers about installing and using this skill.
To build a vector knowledge base, this Skill indexes project files, database metadata, and validated SQL corpora. It dual-routes content into semantic models, metrics, reference SQL, and knowledge stores to provide domain-aware context for retrieval-augmented generation.
Improving LLM-generated SQL accuracy requires indexing schema metadata, business metrics, and reusable query examples. This Skill builds an evolvable, domain-aware context that supplies these specific elements to retrieval-augmented generation workflows.
Yes, you can scope knowledge generation to specific files, tables, datasources, or business domains. The Skill delegates parallel exploration to subagents per domain, classifying content into the correct storage types for focused retrieval.
Indexing validated SQL query corpora is handled by enumerating every natural language question and SQL pair. Schema-non-inferable rules are dual-routed to knowledge stores, requiring no manual curation, enabling analysts to retrieve accurate queries instantly.
No manual curation is required for validated SQL corpora. The Skill automatically enumerates query pairs and applies business domain classification to schema metadata and documentation, building an evolvable context without ongoing manual maintenance.
Use a vector-indexed knowledge store when LLM-generated SQL needs accurate, domain-aware context. It is essential for data engineering workflows where schema metadata, business metrics, and reusable query examples must be retrieved for retrieval-augmented generation.
InnerScopeHearingAnalyze SQL query execution plans and implement indexing strategies.
peterlianpiCreate document embeddings and index them in a vector store for retrieval-augmented generation.
AskTinNguyenDiscovers and analyzes GitHub repositories to extract patterns and knowledge.
LLM-Dev-OpsPerform semantic vector search over large document collections with AgentDB.
