The npm dev dependency @sveltejs/kit depends on @sveltejs/vite-plugin-svelte in the frontend project.

The npm dev dependency @sveltejs/kit depends on @sveltejs/vite-plugin-svelte in the frontend project.

Creates prompt for LLM to generate SQL, including schema context and examples.

Sends prompt to Ollama API for model inference, retrieves generated text.

Internal function to classify questions as structured, textual, or hybrid, using LLM or keywords.

Falls back to regex keyword patterns to classify questions.

Classifies questions using an LLM call for more accurate determination.

Combines structured SQL and semantic search results for complex queries.

Executes SQL queries generated from questions on DuckDB for structured data retrieval.

Performs semantic search via Qdrant and synthesizes answers for unstructured data questions.

The function _find_numeric_columns was fixed to support decimal types by importing Decimal and updating type detection, deployed in commit 507c94b.

Checks if the project has document vectors stored in Qdrant, indicating available documents.

Checks if the project database contains tables, indicating structured data.

Extracts SQL and reasoning from Ollama response, with fallback parsing.

Agent Skills Integration modifies _run_query to generate findings from query results and stream them as insights. The process_message function is expected to call the _run_query function to execute queries, but current data flow problem stops execution before _run_query is reached. In agent_skills.py, the process_message() function calls _run_query asynchronously to generate query results. _run_query method is modified as part of Agent Skills Integration to generate findings from query results _run_query generates findings streamed as 'insight' messages, each rendered as a separate IronClawMessage The process_message() method in agent_skills.py calls the _run_query() function asynchronously to execute queries.

Admin User was configured with language set to Danish to receive all summaries and UI in Danish. The Admin User belongs to the organization Exerun. Danish Language Support was implemented so that Admin User interacts with the system in Danish language.

Playwright E2E tests use the test user admin@exerun.com for authentication and functional validation. The DataLens project is approved and accessed by the user admin@exerun.com.

DataLens requires adding AfterQuery hooks to intercept requests for cost tracking and filtering.

An entity that routes and manages data analysis queries, supporting multiple paths like structured, textual, or hybrid.

User uses the Agent Chat interface to interact with the SVGV budget analysis system Agent Chat integrates with OpenClaw Gateway for processing user queries with Claude Agent Chat interface handles Danish language budget queries from users

Agent Chat interface uses OpenClawHttpClient component for communication with agent backend FindingsPanelNew replaces FindingsBoard in the agent page user interface The Analysis chat interface is a planned feature to provide conversational query capability in the DataLens Platform.

User testing to verify agent chat system's responsiveness and stability, including streaming responses, Danish language support, and no timeout errors, conducted at datalens.exerun.com.

Agent config files (SOUL.md, AGENTS.md, MEMORY.md) are part of the DataLens platform backend. The DataLens platform backend includes agent configuration files such as SOUL.md, AGENTS.md, and MEMORY.md. Agent configuration includes the SOUL.md file. Agent configuration includes the AGENTS.md file. Agent configuration includes the MEMORY.md file. The DataLens platform backend includes an Agent config.

The Agent Gateway module in FastAPI acts as a bridge and uses IronClaw Service for agent session management and skill execution. Agent Gateway manages sessions and multi-tenant context by interacting with PostgreSQL where metadata and agent tables reside. Agent Gateway interacts with Redis as part of the backend ecosystem for caching and background jobs. Agent Gateway depends on the IronClaw Service to handle reasoning loops, skill execution, and memory management via HTTP and WebSocket communication. Agent Gateway is implemented as a FastAPI module that bridges the frontend and IronClaw Service. Agent Gateway accesses PostgreSQL to handle metadata and agent session tables for multi-tenant context. Agent Gateway integrates with existing services including DuckDB for query execution. Agent Gateway integrates with existing Qdrant vector database service for data operations. Agent Gateway integrates with Anthropic API to provide cloud LLM model backend services.

Agent Integration Tests validate the end-to-end flow of DataLens Agent Mode including session lifecycle and model routing.