List all accessible files (owned + shared with you). Returns file metadata including name, type, size, owner, timestamps, and embedding status.

Parameters

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "list_files",
    "arguments": {}
  }
}

Example Response

[
  {
    "id": 42,
    "name": "meeting-notes.pdf",
    "type": "application/pdf",
    "size": 184320,
    "owner": "alice",
    "created": "2026-02-10T14:30:00Z",
    "updated": "2026-02-10T14:30:00Z",
    "embedding_status": "complete",
    "shared": false
  }
]

Read the content of a specific file. Returns extracted text for documents and PDFs, or base64-encoded data for images.

Parameters

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "read_file",
    "arguments": { "file_id": 42 }
  }
}

Example Response

{
  "content": [
    {
      "type": "text",
      "text": "File: meeting-notes.pdf\\n\\nQ1 Planning Meeting\\n..."
    }
  ]
}

Upload a file to the knowledge base. The content must be base64-encoded. Returns a file_id that can be used with get_upload_status to track processing. Supports images, PDFs, text, code files, Office documents, and more.

Parameters

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "upload_file",
    "arguments": {
      "filename": "notes.md",
      "content": "IyBNZWV0aW5nIE5vdGVzCi0gRGlzY3Vzc2VkIFExIHRhcmdldHM="
    }
  }
}

Example Response

{
  "file_id": 87,
  "filename": "notes.md",
  "size_bytes": 42,
  "mime_type": "text/markdown",
  "status": "processing",
  "message": "File uploaded successfully. Use get_upload_status with the file_id to check when processing is complete."
}

Check the processing status of an uploaded file. Use this after upload_file to know when the file is fully processed and searchable. Status values: pending, processing, done, or error.

Parameters

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "get_upload_status",
    "arguments": { "file_id": 87 }
  }
}

Example Response

{
  "file_id": 87,
  "filename": "notes.md",
  "size_bytes": 42,
  "mime_type": "text/markdown",
  "status": "done",
  "created_at": "2026-03-15T01:15:00Z",
  "updated_at": "2026-03-15T01:15:03Z"
}

Semantically search across all text content in the knowledge base. Uses vector embeddings to find results by meaning, not just keywords.

Parameters

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "search_text",
    "arguments": {
      "query": "quarterly revenue projections",
      "limit": 3
    }
  }
}

Example Response

[1] File: q1-report.pdf (ID: 42)
    Owner: alice | Uploaded: 2026-01-15 | Size: 180KB
    Similarity: 92.3%
    Content: Revenue projections for Q1 show a 15% increase...

[2] File: finance-notes.txt (ID: 38)
    Owner: alice | Uploaded: 2026-01-10 | Size: 12KB
    Similarity: 84.7%
    Content: Updated revenue targets based on market analysis...

Search images using a text description (text→image) or by similarity to another image (reverse image search). Powered by CLIP embeddings.

Parameters

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "search_images",
    "arguments": {
      "query": "architecture diagram with microservices"
    }
  }
}

Example Response

[1] File: system-arch.png (ID: 55)
    Owner: bob | Uploaded: 2026-02-01 | Size: 340KB
    Similarity: 78%

[2] File: infra-overview.jpg (ID: 61)
    Owner: alice | Uploaded: 2026-01-28 | Size: 215KB
    Similarity: 65%

List all collections the user has. Collections are organizational folders for grouping related files together.

Parameters

This tool takes no parameters.

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "list_collections",
    "arguments": {}
  }
}

Example Response

[
  {
    "id": 1,
    "name": "Project Alpha",
    "description": "All docs related to Project Alpha",
    "file_count": 12,
    "is_default": false
  },
  {
    "id": 2,
    "name": "Research Papers",
    "description": null,
    "file_count": 8,
    "is_default": false
  }
]

List all files within a specific collection. Accepts either the collection name (case-insensitive) or its numeric ID.

Parameters

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "list_collection_files",
    "arguments": { "collection": "Project Alpha" }
  }
}

Example Response

{
  "collection": {
    "id": 1,
    "name": "Project Alpha",
    "description": "All docs related to Project Alpha"
  },
  "files": [
    {
      "id": 42,
      "name": "spec.pdf",
      "type": "application/pdf",
      "size": 184320,
      "uploaded": "2026-02-10T14:30:00Z"
    }
  ]
}

Semantically search within a specific collection. Same as search_text but scoped to a single collection.

Parameters

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "search_in_collection",
    "arguments": {
      "collection_id": 1,
      "query": "API rate limits",
      "limit": 3
    }
  }
}

Example Response

[1] File: api-docs.md (ID: 44)
    Owner: alice | Uploaded: 2026-02-05 | Size: 8KB
    Similarity: 91.2%
    Content: Rate limiting is set to 100 requests per minute...

Save a memory or note for later retrieval. The content is automatically embedded for semantic search.

Parameters

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "save_memory",
    "arguments": {
      "content": "User prefers TypeScript over JavaScript for all new projects",
      "metadata": { "source": "conversation", "topic": "preferences" }
    }
  }
}

Example Response

Memory saved (ID: 17)

Search through stored memories by semantic similarity. Returns the most relevant memories ranked by match.

Parameters

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "search_memory",
    "arguments": {
      "query": "programming language preferences"
    }
  }
}

Example Response

[1] Memory (ID: 17)
    Created: 2026-02-15 | Source: conversation
    Similarity: 94.1%
    Content: User prefers TypeScript over JavaScript for all new projects

Get storage usage and account tier information.

Parameters

This tool takes no parameters.

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "get_usage",
    "arguments": {}
  }
}

Example Response

{
  "tier": "free",
  "storage_used_mb": 142,
  "total_files": 28,
  "total_memories": 15
}

Get detailed metadata for a specific file, including owner info, timestamps, size, MIME type, embedding status, and collections.

Parameters

Example Request

{
  "method": "tools/call",
  "params": {
    "name": "get_file_metadata",
    "arguments": { "file_id": 42 }
  }
}

Example Response

{
  "id": 42,
  "name": "meeting-notes.pdf",
  "mime_type": "application/pdf",
  "size_bytes": 184320,
  "size_human": "180KB",
  "owner": "alice",
  "owner_email": "[email protected]",
  "created_at": "2026-02-10T14:30:00Z",
  "updated_at": "2026-02-10T14:30:00Z",
  "last_modified_by": "alice",
  "embedding_status": "complete",
  "collections": [
    { "id": 1, "name": "Project Alpha" }
  ]
}