Documentation Index
Fetch the complete documentation index at: /gwenflow/llms.txt
Use this file to discover all available pages before exploring further.
Runnable scripts in the repository, by theme.
examples/ directory of the repository holds runnable scripts. From a clone of the repository:
uv sync --all-groups --frozen
uv pip install yfinance # the finance examples
echo "OPENAI_API_KEY=sk-..." >> .env # or ANTHROPIC_API_KEY, depending on the example
uv run python examples/agent_fx.py
| Script | Shows |
|---|---|
agent_fx.py | a function tool with @agent.tool: exchange rates |
agent_bikes.py | a function tool over a public API, JSON answers |
agent_wikipedia.py | the built-in WikipediaTool |
agent_web.py, agent_websearch_seo.py, agent_websearch_renovation_strategy.py | Tavily search and website reading |
agent_news.py | news aggregation |
agent_structured_output.py | a Pydantic response_model |
agent_async.py | arun and async streaming |
agent_tool_approval.py | a tool that waits for the user’s approval |
tools_app.py | a catalogue of the built-in tools |
| Script | Shows |
|---|---|
agent_finance.py | live market questions with the Yahoo Finance tools |
agent_finance_portfolio.py | a portfolio snapshot as structured output |
agent_finance_research_team.py | a research team: an orchestrator and specialist agents |
agent_finance_screener.py | a stock screener with streaming |
agent_finance_risk.py | a risk memo, structured and tool-heavy |
| Script | Shows |
|---|---|
agent_memory.py | a custom ChatMemoryBuffer as the agent’s history |
agent_long_term_memory.py | long-term memory: the toolset, the consolidator and the curator |
| Script | Shows |
|---|---|
flow_sequential.py | two agents in a row, in plain code |
flow_yaml.py | a YAML flow with FlowRunner (flows/news_summary.yaml) |
flow_temporal.py | the same flow run durably on Temporal, then scheduled every day |
agent_temporal.py | a durable agent, the simplest setup |
agent_temporal_human_approval.py | a durable agent whose tool waits for a human approval, with AgentWorkflow |
pip install "gwenflow[temporal]" and a server: temporal server start-dev.
| Script | Shows |
|---|---|
telemetry.py | OpenTelemetry tracing with a session |
examples/README.md are illustrative: prices, headlines and figures reflect live data at the time you run them.