google-workspace-unlimited
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| LOG_LEVEL | No | Logging level (DEBUG, INFO, WARNING, ERROR) | INFO |
| QDRANT_KEY | No | Qdrant API key (use NONE for no auth) | NONE |
| QDRANT_URL | No | Qdrant vector database URL | http://localhost:6333 |
| SERVER_HOST | No | Server bind address | localhost |
| SERVER_PORT | No | Server port | 8002 |
| ENABLE_HTTPS | No | Enable HTTPS/SSL | false |
| SSL_KEY_FILE | No | Path to SSL private key (Required if ENABLE_HTTPS is true) | |
| FASTMCP_CLOUD | No | Enable cloud deployment mode | false |
| MCP_TRANSPORT | No | Communication protocol (stdio or http). Glama's mcp-proxy expects stdio. | stdio |
| MINIMAL_TOOLS | No | Skip heavy module-wrapper indexing at startup (818-component email types, card_framework widgets). Set to 'true' for fast cold-start in environments with strict startup timeouts (Glama sandbox, CI smoke tests). | false |
| SSL_CERT_FILE | No | Path to SSL certificate (Required if ENABLE_HTTPS is true) | |
| SAMPLING_TOOLS | No | Enable sampling middleware tools (validation agents). Requires VENICE_INFERENCE_KEY, LITELLM_API_KEY, or ANTHROPIC_API_KEY. | false |
| CREDENTIALS_DIR | No | Directory for stored credentials | ./credentials |
| ENABLE_CODE_MODE | No | Enable Code Mode (sandboxed Python tool-call chaining via the `execute` meta-tool) | false |
| GMAIL_ALLOW_LIST | No | Comma-separated allowed email addresses | |
| GOOGLE_CLIENT_ID | No | OAuth 2.0 client ID from Google Cloud Console | |
| OAUTH_REDIRECT_URI | No | Must match Google Console exactly | http://localhost:8002/oauth2callback |
| QDRANT_AUTO_LAUNCH | No | Automatically launch Qdrant via Docker if not reachable. Set to 'false' in single-container environments where Docker-in-Docker isn't available (Glama sandbox). | true |
| SESSION_SECRET_KEY | No | Secret key for session management | |
| ATTACHMENT_TEMP_DIR | No | Temp directory for outbound attachment downloads (Gmail signed-URL serving) | /tmp/gw-mcp-attachments |
| ENABLE_AUDIT_LOGGING | No | Enable security event tracking | |
| GOOGLE_CLIENT_SECRET | No | OAuth 2.0 client secret from Google Cloud Console | |
| DRIVE_UPLOAD_TEMP_DIR | No | Temp directory for staged client→server uploads (only used when DRIVE_UPLOAD_CLIENT_FS=true) | /tmp/gw-mcp-drive-uploads |
| DRIVE_UPLOAD_CLIENT_FS | No | Treat upload_to_drive 'path' as the client's filesystem and orchestrate client→server→Drive transport via signed PUT URL. Set to 'true' on remote/hosted deployments where client and server are on different machines. | false |
| CREDENTIAL_STORAGE_MODE | No | Credential storage mode (FILE_ENCRYPTED, FILE_PLAINTEXT, MEMORY_ONLY) | FILE_ENCRYPTED |
| SESSION_TIMEOUT_MINUTES | No | Session idle timeout in minutes | 60 |
| DRIVE_UPLOAD_MAX_SIZE_MB | No | Maximum size in MB for staged client→server uploads (only used when DRIVE_UPLOAD_CLIENT_FS=true) | 100 |
| DRIVE_UPLOAD_TTL_SECONDS | No | TTL in seconds for staged client→server upload URLs (one-time-use) | 900 |
| GOOGLE_CLIENT_SECRETS_FILE | No | Alternative: path to OAuth JSON file |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
Google Workspace MCP Server - Comprehensive access to Google services.
Authentication
Call
start_google_authwith your email to begin OAuth flowComplete authentication in browser
Call
check_drive_authto verify credentials
Available Services
📁 Google Drive: Cloud storage and file synchronization service
📧 Gmail: Email service with powerful search, filtering, and organization features
📅 Google Calendar: Time management and scheduling service
📄 Google Docs: Document creation and collaboration service
📊 Google Sheets: Spreadsheet and data analysis service
💬 Google Chat: Team messaging and collaboration platform
📝 Google Forms: Survey and form creation service
🎯 Google Slides: Presentation creation and sharing service
📷 Google Photos: Photo and video storage service
👤 Google People API: User profile and contact information service
Tool Management
Use manage_tools to list, enable, or disable tools at runtime.
🗄️ Vector Database
Qdrant: Offline (analytics unavailable)
Instructions updated: 2026-11-13 17:42 UTC
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| logging | {} |
| prompts | {
"listChanged": true
} |
| resources | {
"subscribe": false,
"listChanged": true
} |
| completions | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| tagsA | List this server's tool tags (service areas like gmail, drive, docs, photos) with tool counts. Use when: browsing what capability areas exist before a targeted lookup. For keyword lookup use search; for parameters of known tools use get_schema. Returns: '- tag (N tools)' lines at detail='brief' (default), or every tool listed under each tag at detail='full'. |
| searchA | Find this server's Google Workspace tools by keyword (BM25 over names, descriptions, and tags). Use when: you don't know the exact tool name yet. To browse by category use tags; once you have names, use get_schema for parameters, then execute to call them. To search past results rather than tools, use semantic_search. Returns: matching tools as names + descriptions ('brief', default), parameter markdown ('detailed'), or complete JSON definitions ('full'). An empty result means no keyword match — retry with different terms. |
| get_schemaA | Get parameter schemas for named tools before calling them via execute. Use when: you already have tool names (from search or tags) and need exact parameters. Not for discovery — use search for that. Returns: per-tool parameter markdown ('detailed', default), names + descriptions ('brief'), or full JSON schemas ('full'); unknown names are reported under 'Tools not found'. |
| semantic_searchA | Search the Qdrant vector store of this server's past tool responses and card templates. Use when: looking up previous results, usage history, or analytics ('service:gmail last week', 'tool:search recent', 'overview'). For discovering tools to call, use search; to preview one stored document, pass its point ID to fetch_document. Modes: semantic text similarity; service/tool history queries; analytics ('overview', 'usage stats'); filter_dsl for structured filters; positive/negative point IDs for recommendation; query_dsl for fusion/discover/order-by; prefetch_dsl for multi-stage retrieval. Behavior: read-only against the local Qdrant instance; no Google APIs are called. Returns: scored rows 'score service/tool timestamp id:<point_id>'. Errors: 'Search failed' when Qdrant is unreachable; 'No results' when nothing clears score_threshold (default 0.3 — lower it to widen the net). DSL filter symbols: D_5=DatetimeRange, ʄ=FieldCondition, ƒ=Filter, F_6=FilterSelector, ℏ=HasIdCondition, I_2=IsEmptyCondition, I_0=IsNullCondition, ɱ=MatchAny, ṁ=MatchText, M_0=MatchTextAny, ☆=MatchValue, ř=Range Query-DSL symbols: Å=AcornSearchParams, C_14=ContextExamplePair, ¢=ContextPair, C_0=ContextQuery, D_6=DiscoverInput, D_2=DiscoverQuery, D_4=DiscoverRequest, D_21=DiscoverRequestBatch, ℱ=Fusion, φ=FusionQuery, ø=OrderBy, ɵ=OrderByQuery, ¶=Prefetch, ʔ=QuantizationSearchParams, R_12=RecommendGroupsRequest, R_2=RecommendInput, R_4=RecommendQuery, R_5=RecommendRequest, R_18=RecommendRequestBatch, R_10=RecommendStrategy, ♦=SearchParams Example: ƒ{must=[ʄ{key="tool_name", match=☆{value="send_dynamic_card"}}]} MatchAny: ƒ{must=[ʄ{key="tool_name", match=ɱ{any=["tool_a", "tool_b"]}}]} |
| fetch_documentA | Preview one stored tool response by its Qdrant point ID. Use when: inspecting a hit returned by semantic_search. To find
point IDs in the first place, use semantic_search; for the full
untruncated content, call the Behavior: read-only. Returns: tool name, service, timestamp, user, argument count, and the first 500 characters of stored content. Errors: 'Document not found' for unknown or expired point IDs. |
| tool_activityA | Show usage analytics for this server's tools: call counts, error rates, last-used times. Use when: answering 'what has been used or failing lately'. To read an individual response, pass a sample point ID to fetch_document; to discover tools to call, use search instead. Behavior: read-only aggregation over the Qdrant response store. Returns: a text dashboard grouped by tool_name or user_email, with sample point IDs per group. Errors: 'Analytics failed' when the response store is unreachable. |
| executeA | Run sandboxed Python that calls this server's Google Workspace tools via SANDBOX RESTRICTIONS — these produce SandboxError, avoid them:
Built-in helpers (
|
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| quick_email_demo | Simple: Zero-config Gmail demo with Jinja2 templates and resource integration |
| structured_tool_showcase | Showcase a specific tool with structured response example (random if tool not specified) |
| structured_response_comparison | Compare multiple tools showing original vs structured response formats |
| structured_interactive_demo | Interactive challenge featuring a random tool with structured responses |
| structured_response_benefits | Comprehensive overview of structured response system benefits and implementation |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| manage_tools_dashboard | Read-only dashboard showing tool enable/disable state |
| code_mode_prefab_renderer | Prefab renderer used to draw app cards returned via execute |
| latest_data_dashboard | Dashboard for the most recently called list tool |
| Current User Email | Slim 'most recent authenticated email for this session' — returns email + session_id + authenticated flag. For the full profile (credentials, scopes, etc.) use user://current/profile. |
| Current User Identity | Lightweight active-account identity: email, session, and auth provenance. Optimized for client subscriptions — AuthMiddleware fires notifications/resources/updated for this URI when the active Google account changes between tool calls. For credential details use user://current/profile. |
| Current User Profile | Canonical current-user resource. Superset of user://current/email + user://current/identity + prior profile: email, session_id, authenticated, auth_provenance, identity_source, linked_accounts, and nested auth_status (credential validity, scopes, expiry). Subscription target. |
| Current Authentication Session | Detailed information about the current authentication session including token status, expiration times, and granted scopes |
| Active Authentication Sessions | Administrative view of all active authentication sessions with session details, user information, and expiration status for multi-user session management |
| Recent Gmail Messages Cache | Cached list of recent Gmail messages from the last 7 days with sender information, subjects, snippets, and thread IDs - refreshed every 5 minutes for efficient email monitoring |
| Complete Tools Directory | Comprehensive catalog of all available tools organized by category including Drive, Gmail, Calendar, Chat, Forms, Docs, Sheets, and authentication tools with detailed capability descriptions |
| Detailed Tools Collection | Curated list of detailed tools that use automatic resource templating - no user_google_email parameters required, seamless authentication through OAuth session context |
| Gmail Allow List | Get the configured Gmail allow list for send_gmail_message tool - recipients on this list skip elicitation confirmation |
| All Recent Workspace Items | Get recent items from all Google Workspace services (Drive, Docs, Sheets, Slides, Forms, Photos) in a unified view. |
| Available Template Macros | Get all available Jinja2 macros with usage examples. The EnhancedTemplateMiddleware scans middleware/templates/*.j2 files and extracts: - Macro definitions using {% macro name(...) %} - Usage examples from {# MACRO USAGE EXAMPLE: ... #} comments - Template file paths and descriptions Returns comprehensive macro information including: - Macro names and parameters - Template file locations - Ready-to-use usage examples - Discovery timestamps This enables easy discovery and correct usage of available template macros. Example response: { "macros": { "render_gmail_labels_chips": { "name": "render_gmail_labels_chips", "template_file": "email_card.j2", "usage_example": "{{ render_gmail_labels_chips(service://gmail/labels, 'Label summary') }}" } }, "total_count": 5 } This resource is handled by EnhancedTemplateMiddleware. |
| User Email Template | Simple template resource that returns just the user email string - the most basic resource for tools that need only the email address |
| Qdrant Collections List | List all available Qdrant collections with detailed information including point counts, vector statistics, and status |
| Qdrant Tool Response Cache | Get cached tool responses from Qdrant organized by tool name with point metadata including IDs, timestamps, and user emails |
| Qdrant Middleware Status | Check the status of Qdrant middleware integration including client availability, initialization state, and resource handler status |
TDQS
Scored across 7 tools
The tools are mostly separable: tags and search provide different discovery paths, and semantic_search is clearly scoped to past responses. The main ambiguity is that search, semantic_search, and tags all surface tool/response names in slightly different ways, though the descriptions mitigate this by stating when to use each.
All names use lowercase with underscores, but the structural conventions are mixed: standalone verbs (search, execute), nouns (tags, tool_activity), and verb_noun compounds (get_schema, fetch_document) coexist. This is readable, but there is no single consistent naming pattern across the tool set.
Seven tools is an appropriate, well-scoped count for a discovery, schema, execution, history, and analytics server. Each tool carries a distinct responsibility, with no obvious redundancy or bloat.
The tool surface covers the core lifecycle of discovering tools, fetching schemas, executing calls, and retrieving historical results/analytics. Minor gaps exist, such as no dedicated tool for deleting/cleaning stored history or batch management of point IDs, but these are workable around.