skills.rest
Datus-ai
Datus-aiOFFICIAL
Explore all skills by this author →
WebsiteGitHub

storage-classify

Classify content artifacts into persistent storage tiers by type, lifetime, and reuse pattern.

1.5k|222|Updated Jul 4, 2025
Search Tags:#knowledge management#data persistence#artifact routing#subagent delegation#semantic models#sql indexing#storage classification
One-click install
npx skills add https://github.com/Datus-ai/Datus-agent --skill storage-classify
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: storage-classify
Source: https://github.com/Datus-ai/Datus-agent/tree/main/datus/resources/skills/storage-classify
Command: npx skills add https://github.com/Datus-ai/Datus-agent --skill storage-classify

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Without consistent routing rules, data artifacts like validated SQL, business facts, and metric definitions end up in inconsistent or incorrect storage, making them hard to find, breaking validation rules, and reducing reusability across data projects.

Core Features & Use Cases

  • Intelligent Artifact Classification: Automatically categorizes content into 7 distinct persistent stores (semantic_models, metrics, reference_sql, knowledge, memory, AGENTS.md, skills) based on its type, lifetime, and reuse pattern.
  • Strict Routing Enforcement: Follows project-defined routing rules, such as delegating semantic model generation to dedicated subagents instead of hand-writing YAML, and ensuring one SQL query per reference_sql entry for accurate indexing.
  • Ambiguity Handling: Prompts the user for clarification when content does not clearly fit a store or routing would cause destructive overwrites, preventing accidental data loss. Use Case: When a data engineer finishes validating a new sales metric SQL query, this skill routes it to the reference_sql store via the gen_sql_summary subagent, passes the associated business context to ensure the query is searchable for future reuse, and indexes it correctly in the vector store.

Quick Start

Use the storage-classify skill to route the newly validated customer churn metric definition to the correct persistent storage location.

Frequently Asked Questions about storage-classify

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automatically classify and route data artifacts to persistent storage?▼

Automatically classifying and routing data artifacts to persistent storage involves categorizing content by type, lifetime, and reuse pattern into distinct stores like semantic_models, metrics, and reference_sql to ensure consistent retrieval. It enforces routing rules and context handoff to subagents.

What is the best way to store validated SQL queries for future reuse in data engineering?▼

Storing validated SQL queries for reuse requires routing each query to a dedicated reference_sql store, ensuring one query per entry for accurate indexing. This skill delegates context handoff to subagents, passing business context so the query is searchable and consistently retrievable.

How do you handle ambiguous data artifacts that do not clearly fit a persistent storage tier?▼

Handling ambiguous data artifacts that do not clearly fit a persistent storage tier requires prompting the user for clarification before routing. This prevents destructive overwrites and accidental data loss when content does not match predefined routing rules for stores like knowledge or memory.

Can I use subagent delegation for semantic model generation instead of hand-writing YAML?▼

Subagent delegation for semantic model generation is supported by enforcing strict routing rules that delegate the task to dedicated subagents instead of hand-writing YAML. This ensures semantic models are stored consistently in their correct persistent storage tier.

What types of persistent storage tiers are available for knowledge management and artifact routing?▼

Persistent storage tiers available for knowledge management and artifact routing include seven distinct stores: semantic_models, metrics, reference_sql, knowledge, memory, AGENTS.md, and skills. Each store categorizes content based on its specific type, lifetime, and reuse pattern.

Related Skills

Explore More Data & Analytics Skills→
lucaslessachaveslucaslessachaves
1

databricks-metric-views

Define governed business metrics in YAML with dimensions and measures.

Community
Advanced
markus41markus41
1

vectordb

Manage vector embeddings for semantic search across Pinecone, Weaviate, Chroma, and pgvector.

Community
Intermediate
niitsniits

using-dbt-for-analytics-engineering

Manage dbt data models with SQL transformations, tests, and debugging.

Community
Advanced
mindmorassmindmorass
2

qdrant-patterns

Store and retrieve documents in Qdrant for semantic search.

Community
Advanced