feat: FastAPI Refactor; HITL HTTP; Distributed Workflow Observability

docs/source/components/agents/auto-memory-wrapper/auto-memory-wrapper.md

@@ -199,7 +199,7 @@ async def handle_request(request):
 For quick testing without custom middleware:
 
 ```bash
-curl -X POST http://localhost:8000/chat \
+curl -X POST http://localhost:8000/v1/chat/completions \
   -H "Content-Type: application/json" \
   -H "X-User-ID: test_user_123" \
   -H "conversation-id: test_conv_001" \

docs/source/components/auth/mcp-auth/index.md

@@ -266,15 +266,15 @@ nat serve --config_file examples/MCP/simple_auth_mcp/configs/config-mcp-auth-jir
 
 User Alice:
 ```bash
-curl -X POST http://localhost:8000/generate \
+curl -X POST http://localhost:8000/v1/workflow \
   -H "Content-Type: application/json" \
   -H "Cookie: nat-session=user-alice" \
   -d '{"messages": [{"role": "user", "content": "What is status of AIQ-2342?"}]}'
 ```
 
 User Bob (has a separate MCP client instance):
 ```bash
-curl -X POST http://localhost:8000/generate \
+curl -X POST http://localhost:8000/v1/workflow \
   -H "Content-Type: application/json" \
   -H "Cookie: nat-session=user-bob" \
   -d '{"messages": [{"role": "user", "content": "What is status of AIQ-2507?"}]}'

docs/source/extend/custom-components/custom-functions/per-user-functions.md

@@ -147,13 +147,13 @@ When using the FastAPI front end with `nat serve`, users are identified by the `
 
 ```bash
 # User "alice" makes a request
-curl -X POST http://localhost:8000/generate \
+curl -X POST http://localhost:8000/v1/workflow \
   -H "Content-Type: application/json" \
   -H "Cookie: nat-session=alice" \
   -d ''{"messages": [{"role": "user", "content": "Hello"}]}''
 
 # User "bob" makes a request (isolated from alice)
-curl -X POST http://localhost:8000/generate \
+curl -X POST http://localhost:8000/v1/workflow \
   -H "Content-Type: application/json" \
   -H "Cookie: nat-session=bob" \
   -d ''{"messages": [{"role": "user", "content": "Hello"}]}''

docs/source/improve-workflows/finetuning/concepts.md

@@ -634,7 +634,7 @@ nat finetune --config_file=configs/finetune.yml \
 
 # Using remote endpoint
 nat finetune --config_file=configs/finetune.yml \
-    --endpoint=http://localhost:8000/generate \
+    --endpoint=http://localhost:8000/v1/workflow \
     --endpoint_timeout=600
 ```
 

docs/source/improve-workflows/optimizer.md

@@ -471,7 +471,7 @@ nat optimize --config_file <path_to_config>
 -   `--config_file`: (Required) Path to the JSON or YAML configuration file for your workflow, for example, `config.yaml`. This file should contain the `optimizer` section as described above.
 -   `--dataset`: (Optional) Path to a JSON file containing the dataset for evaluation, such as `eval_dataset.json`. This will override any dataset path specified in the config file. The dataset should be a list of dictionaries, where each dictionary represents a data point and includes the necessary inputs for your workflow and the ground truth for evaluation.
 -   `--result_json_path`: A `JSONPath` expression to extract the result from the workflow's output. Defaults to `$`.
--   `--endpoint`: If you are running your workflow as a service, you can provide the endpoint URL. For example, `http://localhost:8000/generate`.
+-   `--endpoint`: If you are running your workflow as a service, you can provide the endpoint URL. For example, `http://localhost:8000/v1/workflow`.
 -   `--endpoint_timeout`: The timeout in seconds for requests to the endpoint. Defaults to `300`.
 
 Example:

docs/source/reference/cli.md

@@ -537,7 +537,7 @@ Options:
   --skip_completed_entries    Skip the dataset entries that have a generated
                               answer.
   --endpoint TEXT             Use endpoint for running the workflow. Example:
-                              http://localhost:8000/generate
+                              http://localhost:8000/v1/workflow
   --endpoint_timeout INTEGER  HTTP response timeout in seconds. Only relevant
                               if endpoint is specified.  [default: 300]
   --reps INTEGER              Number of repetitions for the evaluation.
@@ -582,7 +582,7 @@ Options:
                                   example, '$.output' will extract the 'output'
                                   field from the result.  [default: $]
   --endpoint TEXT                 Use endpoint for running the workflow.
-                                  Example: http://localhost:8000/generate
+                                  Example: http://localhost:8000/v1/workflow
   --endpoint_timeout INTEGER      HTTP response timeout in seconds. Only
                                   relevant if endpoint is specified.
                                   [default: 300]
@@ -649,7 +649,7 @@ Use a remote endpoint for workflow execution:
 <!-- path-check-skip-begin -->
 ```bash
 nat finetune --config_file=configs/finetune.yml \
-    --endpoint=http://localhost:8000/generate \
+    --endpoint=http://localhost:8000/v1/workflow \
     --endpoint_timeout=600
 ```
 <!-- path-check-skip-end -->
@@ -678,7 +678,7 @@ Options:
                               '$.output' will extract the 'output' field from
                               the result.  [default: $]
   --endpoint TEXT             Use endpoint for running the workflow. Example:
-                              http://localhost:8000/generate
+                              http://localhost:8000/v1/workflow
   --endpoint_timeout INTEGER  HTTP response timeout in seconds. Only relevant
                               if endpoint is specified.  [default: 300]
   --help                      Show this message and exit.