Skip to content

Add AutoRAG mcp server #103

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Apr 30, 2025
Merged

Conversation

G4brym
Copy link
Collaborator

@G4brym G4brym commented Apr 30, 2025

No description provided.

@G4brym G4brym force-pushed the add-autorag-mcp-server branch from 47a7e65 to 48fc7f5 Compare April 30, 2025 13:19
Copy link
Collaborator

@cmsparks cmsparks left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Just the tool description + response related issues, otherwise lgtm

export function registerAutoRAGTools(agent: AutoRAGMCP) {
agent.server.tool(
'list_rags',
'List Rags (vector stores)',
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think we'd probably want a slightly larger description for all of these for the LLMs

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We could also add server instructions (but that's not 100% necessary):

options: { instructions: BASE_INSTRUCTIONS },

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

added clearer descriptions to all fields


agent.server.tool(
'search',
'Search Rag (vector store)',
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we want to add an option to limit search result lengths (alongside a more limited default)? I used autorag with the docs MCP server, which ran into a lot of context length issues.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

added limit to 5 results in chunk search and 10 result when asking for just an ai response

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

also updated all tools response to be simpler for the llm to understand, removing unused fields and converting the chunk search into a single text instead of being a json with a bunch of fields

@G4brym G4brym force-pushed the add-autorag-mcp-server branch from 48fc7f5 to c10c0f3 Compare April 30, 2025 16:22
@G4brym G4brym force-pushed the add-autorag-mcp-server branch from c10c0f3 to 0dbcd19 Compare April 30, 2025 16:29
@Maximo-Guk
Copy link
Member

Maximo-Guk commented Apr 30, 2025

If you could push these branches up to Cloudflare org from the fork, that would be super great ( so we can get CI to run )

@G4brym
Copy link
Collaborator Author

G4brym commented Apr 30, 2025

If you could push these branches up to Cloudflare org from the fork, that would be super great ( so we can get CI to run )

done, ci is now green

@G4brym
Copy link
Collaborator Author

G4brym commented Apr 30, 2025

can this be merged?

@Maximo-Guk
Copy link
Member

Go for it @G4brym !

@Maximo-Guk Maximo-Guk merged commit 7a5d4e5 into cloudflare:main Apr 30, 2025
6 checks passed
@Maximo-Guk
Copy link
Member

Merged on your behalf

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants