Agents for OpenBB
https://github.com/openbb-finance/agents-for-openbb- Category
- Finance
- Rank
- No. 1379Tools index
Previous survey · No. 1348 ·
- Pricing
- Open Source
- Type
- AGENT
- Builder
- openbb-finance
- GitHub
- 383 stars
- Date
About
Custom AI agents that plug into the OpenBB Workspace for financial research and analysis.
What it does
A collection of runnable reference services showing how an assistant exchanges messages, tool results, citations, files, tables, charts, and other response artifacts with OpenBB Workspace over HTTP and server-sent events.
Why it's ranked here
It is a useful protocol cookbook, not a finished product. The examples cover many Workspace response types and integration patterns with compact Python services. The repository itself points readers seeking a production-oriented implementation toward Agent Rita.
What's good
The examples isolate specific capabilities, making the integration contract easier to learn. They demonstrate streamed text, widget retrieval, citations, PDF extraction, charts, tables, HTML artifacts, feedback capture, prompt rewriting, and MCP tool access. Several examples also limit accumulated tool context to reduce model context pressure.
Tradeoffs
Most services call fixed OpenAI models, so they require matching credentials and provider access. Some outputs use hardcoded demonstration data. Feedback persists to one local JSON file. The HTML example manually bypasses SDK validation, and the project depends heavily on the separate OpenBB AI SDK.
How to use it well
Best for Python developers learning the Workspace agent protocol or prototyping one capability at a time. Start from the nearest example, then replace demonstration data, prompts, persistence, and model choices. Use another reference for production architecture, deployment, and operational maturity.
Technical notes+
pyproject.toml defines a Poetry project with package-mode = false, Python ^3.10, FastAPI, Uvicorn, SSE Starlette, OpenAI, PDFPlumber, HTTPX, and openbb-ai ^2.0.1. 20-financial-prompt-optimizer/main.py exposes /agents.json and streams /v1/query. 35-vanilla-agent-pdf/vanilla_agent_pdf/main.py retrieves widget data, extracts URL or base64 PDF text, and emits citations. 33-vanilla-agent-charts/vanilla_agent_charts/main.py and 34-vanilla-agent-tables/vanilla_agent_tables/main.py emit structured artifacts. 42-vanilla-agent-feedback/vanilla_agent_feedback/main.py adds /v1/feedback and writes feedback.json. 39-vanilla-agent-html-artifacts/vanilla_agent_html/main.py manually constructs HTML artifact SSE events because its stated SDK version lacks validation support. The shown portion of 38-vanilla-agent-mcp-tools/vanilla_agent_mcp_tools/main.py maps supplied MCP schemas into an OpenAI function definition and truncates oversized context.
Observed
- Primary language
- Python, requiring Python 3.10 or newer
- Packaging
- Poetry application project with package mode disabled
- HTTP interface
- FastAPI services expose agent descriptors and query endpoints
- Streaming interface
- Responses use server-sent events through SSE Starlette
- Workspace capabilities
- Examples cover widgets, citations, PDFs, charts, tables, HTML artifacts, feedback, and MCP tools
- Core dependency
- OpenBB AI SDK 2.x is declared as a runtime dependency
Read from README.md, pyproject.toml, 20-financial-prompt-optimizer/main.py, 35-vanilla-agent-pdf/vanilla_agent_pdf/main.py, 33-vanilla-agent-charts/vanilla_agent_charts/main.py, 34-vanilla-agent-tables/vanilla_agent_tables/main.py, 42-vanilla-agent-feedback/vanilla_agent_feedback/main.py, 38-vanilla-agent-mcp-tools/vanilla_agent_mcp_tools/main.py, 39-vanilla-agent-html-artifacts/vanilla_agent_html/main.py.
What it can do
Analyze financial data using AI agents
Financial datasets and research queries → AI-generated financial analysis and insights
Generate automated financial research reports
Company tickers, market data, and research parameters → Comprehensive financial research reports
Process natural language financial queries
Plain English questions about financial markets or companies → Structured financial data and analysis results
Integrate AI agents with OpenBB Workspace tools
OpenBB financial data and workspace configurations → Enhanced financial analysis workflows with AI capabilities
Automate financial data collection and processing
Data source specifications and collection parameters → Cleaned and organized financial datasets
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.