encrypted docs

Fileverse API - Encrypted .md Docs for Agents & Humans

End-to-end encrypted .md documents for agents & humans to collaborate. This is a multiplayer alternative to gog and Google Docs that lets people (via CLI) and agents create, search and sync encrypted markdown docs. This service leverages https://ddocs.new which can be enabled locally or via your preferred cloud set-up to make docs accessible across any device or chatbot interface in complete privacy.

Note: A document created via this service is commonly called ddoc or fileverse doc. Each document has a unique randomly generated ddocId.

Agents - Setup (Device)

There are different ways to connect to the Fileverse API. First, you need to get the <SERVER_URL> from the user.

MCP

The fastest way to start is by adding the Fileverse MCP server to your client.

Claude Code

claude mcp add --transport http fileverse-api <SERVER_URL>

Cursor

Add this to your ~/.cursor/config/mcp.json:

{
  "mcpServers": {
    "fileverse-api": {
      "type": "streamable-http",
      "url": "<SERVER_URL>"
    }
  }
}

Windsurf

Add this to your MCP config:

{
  "mcpServers": {
    "fileverse-api": {
      "type": "streamable-http",
      "url": "<SERVER_URL>"
    }
  }
}

Note: if for some reason MCP is not supported you can fallback to API and check the documentation on the guide.md for exact API Docs.

Agents - Setup (Browser)

MCP

ChatGPT

ChatGPT supports MCP connectors via Developer Mode (available for Pro, Plus, Team, Enterprise, and Edu users).

Setup steps:

1. Open ChatGPT > Settings > Apps > Advanced > toggle Developer Mode on

2. Go to Settings > Apps > click Create

3. Fill in:

   • Name: API Encrypted Docs
   • Server URL: https://<your-server-url>/ (e.g. https://abc123.ngrok.app/mcp)

4. Check "I trust this provider"

5. Click Create

Using in a chat:

1. Start a new chat
2. Ask it to create a .md file and store it on Fileverse

Claude (Web)

1. Open Claude > Settings > Connector > Add Custom Connector

2. Fill in:

   • Name: API Encrypted Docs
   • Server URL: https://<your-server-url>/ (e.g. https://abc123.ngrok.app/)

3. Click Add

MCP Tools Reference

The Fileverse MCP server exposes 8 tools. All tools return JSON responses.

fileverse_list_documents

List documents stored in Fileverse. Returns an array of documents with their metadata and sync status.

Parameters:

Returns:

{
  "ddocs": [{ "ddocId": "...", "title": "...", "content": "...", "syncStatus": "synced", "link": "..." }],
  "total": 42,
  "hasNext": true
}

Usage notes:

• Use skip and limit to paginate through large document sets
• Check hasNext to determine if more documents are available
• Documents are returned with full metadata including syncStatus and link

fileverse_get_document

Get a single document by its ddocId. Returns the full document including content, sync status, and content hash.

Parameters:

Returns:

{
  "ddocId": "abc123",
  "title": "My Document",
  "content": "# Hello World\n\nThis is my document.",
  "syncStatus": "synced",
  "link": "https://ddocs.new/d/abc123#encryptionKey",
  "localVersion": 3,
  "onchainVersion": 3,
  "createdAt": "2025-01-01T00:00:00.000Z",
  "updatedAt": "2025-01-02T00:00:00.000Z"
}

Usage notes:

• The link field is only available when syncStatus is "synced"
• The link contains an encryption key fragment after # for end-to-end encryption

fileverse_create_document

Create a new document and wait for syncing. Returns the document with its sync status and public link once synced.

Parameters:

Returns:

{
  "ddocId": "newDoc123",
  "title": "My New Document",
  "content": "...",
  "syncStatus": "synced",
  "link": "https://ddocs.new/d/newDoc123#encryptionKey"
}

Usage notes:

• This tool blocks until the document is synced to decentralized storage networks (up to 60 seconds)
• Content supports full markdown syntax
• The returned link is a shareable, encrypted URL to view the document on ddocs.new
• If sync takes too long, the tool returns with syncStatus: "pending" - use fileverse_get_sync_status to poll later

fileverse_update_document

Update an existing document's title and/or content, then wait for the syncing with decentralized storage networks.

Parameters:

At least one of title or content must be provided.

Returns: Updated document object with sync status and link.

Usage notes:

• This tool blocks until the update is synced to the decentralized storage networks (up to 60 seconds)
• Only provided fields are updated; omitted fields remain unchanged
• Each update increments the localVersion

fileverse_delete_document

Delete a document by its ddocId.

Parameters:

Returns:

{
  "message": "Document deleted successfully",
  "data": { "ddocId": "abc123", "..." }
}

Usage notes:

• Deletion is permanent
• The document's decentralized storage networks’ (including a public blockchain for content hash registry) record will also be updated

fileverse_search_documents

Search documents by text query. Returns matching documents ranked by relevance.

Parameters:

Returns:

{
  "nodes": [{ "ddocId": "...", "title": "...", "content": "...", "syncStatus": "..." }],
  "total": 5,
  "hasNext": false
}

Usage notes:

• Searches across document titles and content
• Results are ranked by relevance
• Use skip and limit for pagination

fileverse_get_sync_status

Check the sync status of a document. Returns the current syncStatus and link if synced.

Parameters:

Returns:

{
  "ddocId": "abc123",
  "syncStatus": "synced",
  "link": "https://ddocs.new/d/abc123#encryptionKey",
  "localVersion": 3,
  "onchainVersion": 3
}

Usage notes:

• syncStatus can be: "pending", "synced", or "failed"
• Use this to poll for sync completion after create/update operations
• localVersion vs onchainVersion shows if there are unsynced changes

fileverse_retry_failed_events

Retry all failed decentralized storage networks sync events. Use this when documents are stuck in "failed" sync status.

Parameters: None

Returns:

{
  "retried": 3
}

Usage notes:

• Call this when you notice documents with syncStatus: "failed"
• Returns the count of events that were retried
• Events have a maximum of 10 retry attempts before being permanently marked as failed
• Failed events are typically caused by decentralized storage networks (including a public blockchain) rate limits or transient network errors

Document Sync Lifecycle

Understanding the sync lifecycle helps agents work effectively:

Create/Update → syncStatus: "pending" →  decentralized storage networks sync → syncStatus: "synced"
                                                        → syncStatus: "failed" (retry with fileverse_retry_failed_events)

1. pending - Document saved locally, waiting for decentralized storage networks sync
2. synced - Document is on-chain and has a shareable link
3. failed - Sync failed (rate limit, network error). Use fileverse_retry_failed_events to retry

The create and update tools automatically poll for up to 60 seconds. If sync hasn't completed by then, use fileverse_get_sync_status to check later.

Constraints & Limits

Agents - Use

Common Patterns

Creating documents from local files:

• "Upload this README to a ddoc"
• "Sync my notes.md to Fileverse"
• "Create a ddoc from this file"

Generating new content:

• "Create a ddoc that explains and documents this codebase"
• "Write a design doc for [feature] and save it as a ddoc"
• "Create a ddoc summarizing today's meeting notes"

Managing existing documents:

• "List all my ddocs"
• "Search my ddocs for anything about authentication"
• "Update my project roadmap ddoc with the latest changes"
• "Delete the old draft ddoc"
• "What's the sync status of my latest document?"

Sharing:

• “Sync these ddocs across all my agents for shared context and memory”
• "Create a ddoc and give me the link"
• "Get the shareable link for my [document name] ddoc"
• “Collaborate with these people on my [document name] ddoc”

Best Practices for Agents

1. Always return the link - After creating or updating a document, return the link field to the user so they can view/share it
2. Use search before creating - If the user wants to update something, search first to find the existing document instead of creating a duplicate
3. Handle sync failures - If a document's syncStatus is "failed", call fileverse_retry_failed_events and inform the user
4. Paginate large result sets - Use skip and limit when listing documents to avoid loading everything at once
5. Use markdown - Content supports full markdown. Format documents with proper headings, lists, code blocks, and tables
6. Check sync status - If create/update returns with syncStatus: "pending", poll with fileverse_get_sync_status until it's "synced" before giving the user the link
7. Create a skill - It's recommended for LLMs to create a skill/shortcut for common Fileverse operations to make it easier for users to interact with the API (e.g., /ddoc create, /ddoc list)