Example Graphs¶
The demo API registers the following graphs. They demonstrate LGOS features;
none is installed as a built-in model by the langgraph-openai-serve package.
Each registration also supplies a short GraphConfig.description used by the
demo model catalogs.
| Model | Demonstrates | Graph feature | Graph-specific dependency |
|---|---|---|---|
advanced-graph |
General chat, gateway tools, uploaded-file Q&A, routed cited research, and approval before saving a searchable note | client_events, file_inputs, interrupts, mcp_tools |
Responses model, selected gateway, OpenAI-compatible vector service, and PostgreSQL |
custom-input-output-context |
Request, output, and typed runtime-context adapters | None | None |
citation-events |
Structured OpenAI URL citations with portable Markdown content | None | None |
file-input |
Central Files API IDs resolved into OpenAI Responses file inputs | file_inputs |
Central Files API and upstream Responses model |
mcp-mock |
Async MCP-style tool discovery and an agent tool loop | None | None |
mcp-postgres |
Read-only database questions with MCP discovery and execution owned by the native UI client | mcp_tools |
Upstream model, selected gateway, DBHub, and PostgreSQL |
complex-subgraphs |
Router-selected subgraphs, status, and nested streamed output | client_events |
None |
custom-event-showcase |
Filtering internal progress and artifact events at the API boundary | client_events |
None |
multi-node-streaming |
Two sequential fake-model nodes contributing ordered text to one assistant message | None | None |
response-outcomes |
Native refusal content and incomplete terminal responses | None | None |
status-events |
Portable status updates for native client UI | client_events |
None |
persistent-plot-agent |
A tool-calling agent with an editable thread-scoped chart | None | Upstream model, Files API, and PostgreSQL store |
interruptible-approval |
Durable choice-or-text human review before protected actions | interrupts |
PostgreSQL checkpointer and run coordinator |
simple-graph |
Streamed model output and discoverable runtime settings | None | Upstream chat model |
simple-graph-external-tools |
Client-provided function tools returned as model tool calls | None | Upstream chat model |
server-tool |
Installed package versions and OpenAI-compatible web search selected by the client | None | Upstream model plus SearXNG, Degoog, or upstream OpenAI search |
lgos-rag |
Agentic retrieval with structured URL citations over the packaged demo corpus | client_events |
Upstream chat and embedding models |
The demo API opens its PostgreSQL runtime during application startup, so PostgreSQL must be available even when you call a provider-free graph. Start it with the demo API instructions.
persistent-plot-agent stores application data with a LangGraph Store.
interruptible-approval checkpoints graph execution. advanced-graph uses both:
checkpoints for human review and Store receipts for vector-service uploads. None
makes LGOS the owner of UI conversation history.
Source Map¶
All graph code is owned by the independent demo/api project:
demo/api/src/lgos_demo_api/app.pyregisters graph names as OpenAI model names.demo/api/src/lgos_demo_api/graphs/contains every graph and adapter listed above.demo/api/src/lgos_demo_api/corpus/contains the Markdown packaged with thelgos-ragexample.
Continue with Run the Demo API for request examples or Chainlit and Open WebUI for UI behavior.