Adding Context

Context is information you attach to your conversation so the assistant answers with the right environment, data, and priorities in mind. Add context from the @ menu or the Add Context button in the chat sidebar.

Context types

Below are all context types you can attach via the @ menu or the Add Context button in the chat sidebar. These map directly to the options shown in the context picker UI.

Host

• What it is: The database host you want the app to use for actions and insights.
• What it does: Anchors queries, operations, alerts, and DB element lookups to a specific environment.
• When to use: Always set a host before asking questions about data, schema, or operations.

Templates

• What it is: Reusable snippets and guides added to the conversation for additional expertise or structure.
• What it does: Supplies ready-made context like best practices, checklists, or analysis recipes the AI can use.
• When to use: To quickly add known instructions or domain knowledge without retyping them.

System Prompts

• What it is: Global guidance for the AI’s behavior and priorities during this chat.
• What it does: Influences tone, safety, depth, and the way answers are structured.
• When to use: To steer the assistant toward specific objectives or styles (for example, performance tuning focus).

ClickHouse DB Elements

• What it is: Databases, tables, and columns sourced from ClickHouse on the selected host.
• What it does: Attaches precise schema elements so answers can reference the right entities.
• When to use: When asking about ClickHouse schemas, queries, or data issues.

Postgres DB Elements

• What it is: Schemas, tables, and columns from PostgreSQL on the selected host (and database).
• What it does: Gives the AI exact schema context to analyze Postgres-specific questions.
• When to use: When troubleshooting or optimizing queries for Postgres.

MySQL DB Elements

• What it is: Databases, tables, and columns from MySQL on the selected host.
• What it does: Adds the exact MySQL objects you’re discussing to eliminate ambiguity.
• When to use: When exploring or debugging MySQL data models and queries.

Firing Problems

• What it is: Active alerts and issues currently detected on the selected hosts.
• What it does: Brings live problems into context so the AI can explain, triage, or suggest remediation.
• When to use: During incident response or when reviewing ongoing production issues.

Operations

• What it is: Live operations such as running processes, queries, merges, or mutations on the selected hosts.
• What it does: Shares real-time workload details to analyze bottlenecks or anomalies.
• When to use: To diagnose slowdowns or understand what’s currently running.