Dash is a Python-based self-learning data agent that grounds answers using six layers of context and provides interfaces via Slack, terminal, or AgentOS UI. It includes a FastAPI API, Docker-based deployment, and knowledge management for queries and business rules.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
Dash is a self-learning data agent built with systems engineering principles. It grounds answers in 6 layers of context and improves with every query. Access is via Slack, the terminal, or the AgentOS web UI.
How it works
Dash uses a multi-layer knowledge framework:
- Data engineering with six grounded context layers (Table Usage, Human Annotations, Query Patterns, Institutional Knowledge, Learnings, Runtime Context).
- A self-learning loop that retrieves knowledge and learnings, reasons about intent, generates grounded SQL, executes, interprets results, and saves learnings.
- Separate knowledge and learnings stores, with a dual schema enforcement separating public data from agent-managed data.
- Interfaces: REST API (FastAPI), Slack, and AgentOS UI.
Getting started
Quick Start steps from README:
# Clone the repo
git clone https://github.com/agno-agi/dash.git && cd dash
cp example.env .env
# Edit .env and add your OPENAI_API_KEY
# Start the system
docker compose up -d --build
# Generate sample data and load knowledge
docker exec -it dash-api python scripts/generate_data.py
docker exec -it dash-api python scripts/load_knowledge.py
Confirm Dash is running at http://localhost:8000/docs.
Connect to the Web UI
- Open os.agno.com and login
- Add OS → Local →
http://localhost:8000 - Click "Connect"
Recent releases
Releases: latest 0; none listed in the provided data.
Traction
Stars: 2242 (as of the provided data). Forks: 249. Open issues: 14.
Behind the repo
Not included in the provided content beyond deployment options (Railway and Docker) and AgentOS integration details.
Caveats
Environment and deployment details include:
- Production requires a JWT_VERIFICATION_KEY from AgentOS and RBAC with auth when RUNTIME_ENV=prd; local development runs without auth.
- Production endpoints enforce schema-level access controls and read-only analyst mode by default.
Load knowledge and data scripts:
python scripts/load_knowledge.py # Upsert changes
python scripts/load_knowledge.py --recreate # Fresh start