feat: [server] Add Human in the Loop example with FastAPI integration #630

leehuwuj · 2025-05-19T04:37:58Z

Introduced a new example in main.py demonstrating the use of LlamaIndexServer with a human in the loop.
Enhanced chat API to support resuming workflows with human responses.
Implemented WorkflowService for context management in workflow.py.

- Introduced a new example in `main.py` demonstrating the use of LlamaIndexServer with a human in the loop. - Added a README.md file to explain the example. - Enhanced chat API to support resuming workflows with human responses. - Implemented `WorkflowService` for context management in `workflow.py`.

changeset-bot · 2025-05-19T04:38:01Z

⚠️ No Changeset found

Latest commit: 91e9ae1

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

This PR includes no changesets

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

coderabbitai · 2025-05-19T04:38:04Z

Important

Review skipped

Draft detected.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

Review comments: Directly reply to a review comment made by CodeRabbit. Example:
- I pushed a fix in commit <commit_id>, please review it.
- Explain this complex logic.
- Open a follow-up GitHub issue for this discussion.
Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
- @coderabbitai explain this code block.
- @coderabbitai modularize this function.
PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
- @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
- @coderabbitai read src/utils.ts and explain its main purpose.
- @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
- @coderabbitai help me debug CodeRabbit configuration file.

Support

Need help? Create a ticket on our support page for assistance with any issues or questions.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

@coderabbitai pause to pause the reviews on a PR.
@coderabbitai resume to resume the paused reviews.
@coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
@coderabbitai full review to do a full review from scratch and review all the files again.
@coderabbitai summary to regenerate the summary of the PR.
@coderabbitai generate docstrings to generate docstrings for this PR.
@coderabbitai generate sequence diagram to generate a sequence diagram of the changes in this PR.
@coderabbitai resolve resolve all the CodeRabbit review comments.
@coderabbitai configuration to show the current CodeRabbit configuration for the repository.
@coderabbitai help to get help.

Other keywords and placeholders

Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
Add @coderabbitai anywhere in the PR title to generate the title automatically.

Documentation and Community

Visit our Documentation for detailed information on how to use CodeRabbit.
Join our Discord Community to get help, request features, and share feedback.
Follow us on X/Twitter for updates and announcements.

marcusschiesser · 2025-05-19T05:00:35Z

python/llama-index-server/llama_index/server/api/routers/chat.py

@@ -129,6 +179,25 @@ async def _text_stream(
            elif hasattr(event, "to_response"):
                event_response = event.to_response()
                yield VercelStreamResponse.convert_data(event_response)
+            elif isinstance(event, InputRequiredEvent):
+                run_id = handler.workflow_handler.run_id


what is this run_id - how is it guaranteed that it's not the same between requests?

A run_id is generated each time we trigger a workflow run, which is unique for every request.
see: https://github.com/run-llama/llama_index/blob/c09c6ae228e6436a00568ad6b03b582307668b66/llama-index-core/llama_index/core/workflow/workflow.py#L239
So it's ideal for storing the workflow context.

python/llama-index-server/llama_index/server/api/routers/chat.py

marcusschiesser · 2025-05-19T05:03:38Z

python/llama-index-server/llama_index/server/api/utils/workflow.py

+import json
+from pathlib import Path
+
+from llama_index.core.workflow import Context, JsonSerializer, Workflow


good is using https://docs.llamaindex.ai/en/stable/understanding/agent/state/

python/llama-index-server/llama_index/server/api/routers/chat.py

marcusschiesser · 2025-05-19T05:07:07Z

python/llama-index-server/llama_index/server/api/routers/chat.py

 from llama_index.server.services.llamacloud import LlamaCloudFileService


+class ResumeRequest(BaseModel):
+    chat_id: str
+    response: str


use might response with a more complex answer (e.g. an image and a text) - in TS we're using MessageContent

Suggested change

response: str

response: MessageContent

python/llama-index-server/examples/hitl/main.py

marcusschiesser · 2025-05-19T05:26:15Z

python/llama-index-server/llama_index/server/api/routers/chat.py

+            logger.error(e)
+            raise HTTPException(status_code=500, detail=str(e))
+
+    @router.post("/resume")


to ensure this endpoint is working with the UI, it should have the same contract as chat

how about:

class ChatRequest(BaseModel): chat_id: Optional[str] = None messages: List[ChatAPIMessage] data: Optional[Any] = None config: Optional[ChatConfig] = ChatConfig() @field_validator("messages") def validate_messages(cls, v: List[ChatAPIMessage], info: "ValidationInfo") -> List[ChatAPIMessage]: if not v: raise ValueError("Messages list cannot be empty.") chat_id_value = info.data.get("chat_id") if chat_id_value is not None: if len(v) != 1: raise ValueError( "If chat_id is provided, messages list must contain exactly one message." ) if v[-1].role != MessageRole.USER: raise ValueError( "Last message must be from user." ) return v

marcusschiesser · 2025-05-19T05:27:46Z

python/llama-index-server/llama_index/server/api/routers/chat.py

+                    run_id=run_id,
+                    ctx=ctx,
+                )
+                yield VercelStreamResponse.convert_data(


this shouldn't be in the streaming code, just extend to_response of InputRequiredEvent

marcusschiesser · 2025-05-19T05:31:44Z

python/llama-index-server/examples/hitl/main.py

+    question = "Where is your place now?"
+    event = await ctx.wait_for_event(
+        HumanResponseEvent,
+        waiter_id=question,


why is the question an id?

python/llama-index-server/examples/hitl/main.py

leehuwuj · 2025-05-20T08:03:47Z

packages/server/next/app/components/ui/chat/custom/events/human-response.tsx

Just a simple UI to demonstrate FE integration

ok, keep in mind that we can't add workflow specific components to the server package (they must be in the dynamic componentsDir), so this is only for testing

marcusschiesser · 2025-05-21T04:44:06Z

python/llama-index-server/llama_index/server/api/models.py


 class ChatRequest(BaseModel):
+    id: str  # provided by FE


is it the same as https://ai-sdk.dev/docs/reference/ai-sdk-ui/use-chat#id ? when does it update? e.g. if the user reloads the UI page?

marcusschiesser · 2025-05-21T04:46:57Z

packages/server/next/app/components/ui/chat/custom/events/human-response.tsx

ok, keep in mind that we can't add workflow specific components to the server package (they must be in the dynamic componentsDir), so this is only for testing

marcusschiesser · 2025-05-21T05:01:07Z

python/llama-index-server/llama_index/server/api/routers/chat.py

+            human_response = last_message.human_response
+            if human_response:
+                previous_ctx = WorkflowService.load_context(request.id, workflow)
+                # send a new human response event then resume the workflow with the previous context
+                previous_ctx.send_event(
+                    HumanResponseEvent(
+                        response=human_response,
+                    )
+                )


try to separete persistence concern and HITL - if request.id is sent, we should re-create the workflow

Suggested change

human_response = last_message.human_response

if human_response:

previous_ctx = WorkflowService.load_context(request.id, workflow)

# send a new human response event then resume the workflow with the previous context

previous_ctx.send_event(

HumanResponseEvent(

response=human_response,

)

)

if request.id:

previous_ctx = WorkflowService.load_context(request.id, workflow)

# send a new human response event then resume the workflow with the previous context

then we need to run the workflow with a specific event, either start event as is or the human response event

marcusschiesser reviewed May 19, 2025

View reviewed changes

python/llama-index-server/examples/hitl/main.py Outdated Show resolved Hide resolved

marcusschiesser reviewed May 19, 2025

View reviewed changes

python/llama-index-server/examples/hitl/main.py Outdated Show resolved Hide resolved

leehuwuj added 4 commits May 19, 2025 15:19

update usecase

68afe02

better explain

8b0f200

fix: update app title in CLI Assistant example

ee2b6a4

add ui integration

91e9ae1

leehuwuj commented May 20, 2025

View reviewed changes

marcusschiesser reviewed May 21, 2025

View reviewed changes

feat: [server] Add Human in the Loop example with FastAPI integration #630

Are you sure you want to change the base?

feat: [server] Add Human in the Loop example with FastAPI integration #630

Uh oh!

Conversation

leehuwuj commented May 19, 2025

Uh oh!

changeset-bot bot commented May 19, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

⚠️ No Changeset found

Uh oh!

coderabbitai bot commented May 19, 2025

Review skipped

Chat

Support

CodeRabbit Commands (Invoked using PR comments)

Other keywords and placeholders

Documentation and Community

Uh oh!

Choose a reason for hiding this comment

Uh oh!

leehuwuj May 19, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

changeset-bot bot commented May 19, 2025 •

edited

Loading

leehuwuj May 19, 2025 •

edited

Loading