Graphlit MCP Server

Tools Graphlit MCP Server exposes (71)

Listed by the server itself during our scan of commit 1b94b88.

• configureProject

  Configures the default content workflow and conversation specification for the Graphlit project. Only needed if user asks to configure the project defaults. *Do not* call unless specifically asked for by the user. To reset the project configuration to 'factory state', assign False or null to

• queryProjectUsage

  Queries project usage records. Usage record name describes the operation, i.e. 'Prompt completion', 'Text embedding', 'GraphQL', 'Entity Event'. 'GraphQL' usage records are used for GraphQL operations, i.e. 'queryContents', 'retrieveSources', 'askGraphlit', etc. 'Entity Event' usage reco

• askGraphlit

  Ask questions about using the Graphlit Platform, or specifically about the Graphlit API or SDKs. When the user asks about how to use the Graphlit API or SDKs, use this tool to provide a code sample in Python, TypeScript or C#. Accepts an LLM user prompt. Returns the LLM prompt completion

• promptConversation

  Prompts an LLM conversation about your entire Graphlit knowledge base. Uses hybrid vector search based on user prompt for locating relevant content sources. Uses LLM to complete the user prompt with the configured LLM. Maintains conversation history between 'user' and LLM 'assistant'.

• retrieveSources

  Retrieve relevant content sources from Graphlit knowledge base. Do *not* use for retrieving content by content identifier - retrieve content resource instead, with URI 'contents://{id}'. Accepts an LLM user prompt for content retrieval. For best retrieval quality, provide only key words or phras

• retrieveImages

  Retrieve images from Graphlit knowledge base. Provides image-specific retrieval when image similarity search is desired. Do *not* use for retrieving content by content identifier - retrieve content resource instead, with URI 'contents://{id}'. Accepts image URL. Image will be used for simila

• extractText

  Extracts JSON data from text using LLM. Accepts text to be extracted, and JSON schema which describes the data which will be extracted. JSON schema needs be of type 'object' and include 'properties' and 'required' fields. Optionally accepts text prompt which is provided to LLM to guide data

• createCollection

  Create a collection. Accepts a collection name, and optional list of content identifiers to add to collection. Returns the collection identifier

• addContentsToCollection

  Add contents to a collection. Accepts a collection identifier and a list of content identifiers to add to collection. Returns the collection identifier.

• removeContentsFromCollection

  Remove contents from collection. Accepts a collection identifier and a list of content identifiers to remove from collection. Returns the collection identifier.

• deleteContent

  Deletes content from Graphlit knowledge base. Accepts content identifier. Returns the content identifier and content state, i.e. Deleted.

• deleteConversation

  Deletes conversation from Graphlit knowledge base. Accepts conversation identifier. Returns the conversation identifier and content state, i.e. Deleted.

• deleteCollection

  Deletes collection from Graphlit knowledge base. Does *not* delete the contents in the collection, only the collection itself. Accepts collection identifier. Returns the collection identifier and collection state, i.e. Deleted.

• deleteFeed

  Deletes feed from Graphlit knowledge base. *Does* delete the contents in the feed, in addition to the feed itself. Accepts feed identifier. Returns the feed identifier and feed state, i.e. Deleted.

• deleteFeeds

  Deletes feeds from Graphlit knowledge base. *Does* delete the contents in the feed, in addition to the feed itself. Accepts optional feed type filter to limit the feeds which will be deleted. Also accepts optional limit of how many feeds to delete, defaults to 100. Returns the feed i

• deleteCollections

  Deletes collections from Graphlit knowledge base. Does *not* delete the contents in the collections, only the collections themselves. Accepts optional limit of how many collections to delete, defaults to 100. Returns the collection identifiers and collection state, i.e. Deleted.

• deleteConversations

  Deletes conversations from Graphlit knowledge base. Accepts optional limit of how many conversations to delete, defaults to 100. Returns the conversation identifiers and conversation state, i.e. Deleted.

• deleteContents

  Deletes contents from Graphlit knowledge base. Accepts optional content type and file type filters to limit the contents which will be deleted. Also accepts optional limit of how many contents to delete, defaults to 1000. Returns the content identifiers and content state, i.e. Deleted.

• queryContents

  Query contents from Graphlit knowledge base. Do *not* use for retrieving content by content identifier - retrieve content resource instead, with URI 'contents://{id}'. Accepts optional content name, content type and file type for metadata filtering. Accepts optional hybrid vector search quer

• queryCollections

  Query collections from Graphlit knowledge base. Do *not* use for retrieving collection by collection identifier - retrieve collection resource instead, with URI 'collections://{id}'. Accepts optional collection name for metadata filtering. Returns the matching collections, including their co

• queryFeeds

  Query feeds from Graphlit knowledge base. Do *not* use for retrieving feed by feed identifier - retrieve feed resource instead, with URI 'feeds://{id}'. Accepts optional feed name and feed type for metadata filtering. Returns the matching feeds, including their feed resource URI to retrieve

• queryConversations

  Query conversations from Graphlit knowledge base. Do *not* use for retrieving conversation by conversation identifier - retrieve conversation resource instead, with URI 'conversations://{id}'. Accepts optional hybrid vector search query. Accepts optional recency filter (defaults to null, mea

• isContentDone

  Check if content has completed asynchronous ingestion. Accepts a content identifier which was returned from one of the non-feed ingestion tools, like ingestUrl. Returns whether the content is done or not.

• isFeedDone

  Check if an asynchronous feed has completed ingesting all the available content. Accepts a feed identifier which was returned from one of the ingestion tools, like ingestGoogleDriveFiles. Returns whether the feed is done or not.

• listNotionDatabases

  Lists available Notion databases. Requires environment variable to be configured: NOTION_API_KEY. Returns a list of Notion databases, where the database identifier can be used with ingestNotionPages to ingest pages into Graphlit knowledge base.

• listDropboxFolders

  Lists available Dropbox folders. Requires environment variables to be configured: DROPBOX_APP_KEY, DROPBOX_APP_SECRET, DROPBOX_REFRESH_TOKEN. Returns a list of Dropbox folders that can be used with file ingestion tools.

• listBoxFolders

  Lists available Box folders. Requires environment variables to be configured: BOX_CLIENT_ID, BOX_CLIENT_SECRET, BOX_REFRESH_TOKEN. Returns a list of Box folders that can be used with file ingestion tools.

• listDiscordGuilds

  Lists available Discord guilds (servers). Requires environment variable to be configured: DISCORD_BOT_TOKEN. Returns a list of Discord guilds that the bot has access to.

• listDiscordChannels

  Lists available Discord channels in a guild. Requires environment variable to be configured: DISCORD_BOT_TOKEN. Returns a list of Discord channels that can be used with Discord ingestion tools.

• listNotionPages

  Lists pages from a Notion database. Requires environment variable to be configured: NOTION_API_KEY. Returns a list of Notion pages in the specified database.

• listGoogleCalendars

  Lists available Google calendars. Requires environment variables to be configured: GOOGLE_CLIENT_ID, GOOGLE_CLIENT_SECRET, GOOGLE_REFRESH_TOKEN. Returns a list of Google calendars that can be used with calendar ingestion tools.

• listMicrosoftCalendars

  Lists available Microsoft calendars. Requires environment variables to be configured: MICROSOFT_CLIENT_ID, MICROSOFT_CLIENT_SECRET, MICROSOFT_REFRESH_TOKEN. Returns a list of Microsoft calendars that can be used with calendar ingestion tools.

• listLinearProjects

  Lists available Linear projects. Requires environment variable to be configured: LINEAR_API_KEY. Returns a list of Linear projects, where the project name can be used with ingestLinearIssues to ingest issues into Graphlit knowledge base.

• listSlackChannels

  Lists available Slack channels. Requires environment variable to be configured: SLACK_BOT_TOKEN. Returns a list of Slack channels, where the channel name can be used with ingestSlackMessages to ingest messages into Graphlit knowledge base.

• listSharePointLibraries

  Lists available SharePoint libraries. Requires environment variables to be configured: SHAREPOINT_CLIENT_ID, SHAREPOINT_CLIENT_SECRET, SHAREPOINT_REFRESH_TOKEN. Returns a list of SharePoint libraries, where the selected libraryId can be used with listSharePointFolders to enumerate SharePoint

• listSharePointFolders

  Lists available SharePoint folders. Requires environment variables to be configured: SHAREPOINT_CLIENT_ID, SHAREPOINT_CLIENT_SECRET, SHAREPOINT_REFRESH_TOKEN. Returns a list of SharePoint folders, which can be used with ingestSharePointFiles to ingest files into Graphlit knowledge base.

• ingestSharePointFiles

  Ingests files from SharePoint library into Graphlit knowledge base. Accepts a SharePoint libraryId and an optional folderId to ingest files from a specific SharePoint folder. Libraries can be enumerated with listSharePointLibraries and library folders with listSharePointFolders. Requires

• ingestOneDriveFiles

  Ingests files from OneDrive into Graphlit knowledge base. Accepts optional OneDrive folder identifier, and an optional read limit for the number of files to ingest. If no folder identifier provided, ingests files from root OneDrive folder. Requires environment variables to be configured:

• ingestGoogleDriveFiles

  Ingests files from Google Drive into Graphlit knowledge base. Accepts optional Google Drive folder identifier, and an optional read limit for the number of files to ingest. For example, with Google Drive URI (https://drive.google.com/drive/u/0/folders/32tzhRD12KDh2hXABY8OZRFv7Smy8WBkQ), the

• ingestDropboxFiles

  Ingests files from Dropbox into Graphlit knowledge base. Accepts optional relative path to Dropbox folder (i.e. /Pictures), and an optional read limit for the number of files to ingest. If no path provided, ingests files from root Dropbox folder. Requires environment variables to be conf

At a glance

Stars

378

Forks

55

Language

TypeScript

License

MIT

Created

Mar 2025

Last update

Jan 12, 2026

Type

MCP Server

Price

Free

Topics

Questions about Graphlit MCP Server

Guides

• How to Install an MCP Server in Claude, Cursor and VS Code
• MCP Server Security: How to Vet a Server Before You Use It
• MCP Tool Poisoning: Malicious Descriptions and Rug Pulls
• Glossary of MCP, skills and agent security terms
• More Web Scraping & Data agents, MCP servers & skills