> ## Documentation Index > Fetch the complete documentation index at: https://docs.ascendgtm.net/llms.txt > Use this file to discover all available pages before exploring further. # MCP Tools Reference > All 34 tools available via the MCP endpoint with parameters and examples. # MCP Tools Reference All 34 tools available via the MCP endpoint. Each tool accepts JSON parameters and returns structured responses. ## Curated Tools (24) ### google\_ads\_query Execute a Google Ads Query Language (GAQL) query. | Parameter | Type | Required | Description | | ------------- | ------ | -------- | -------------------------------------------------- | | `gaql` | string | Yes | GAQL query string | | `customer_id` | string | No | Google Ads customer ID (defaults to tenant config) | | `account_id` | string | No | Multi-account selector | ### hubspot\_crm Full CRUD on HubSpot CRM objects. | Parameter | Type | Required | Description | | ------------------ | --------- | -------- | ----------------------------------------------------- | | `object_type` | enum | Yes | `contacts`, `companies`, `deals`, `tickets` | | `action` | enum | Yes | `search`, `list`, `get`, `create`, `update`, `delete` | | `filters` | array | No | Search filters (AND combined) | | `properties` | string\[] | No | Fields to return | | `properties_write` | object | No | Field values for create/update | | `id` | string | No | Record ID (for get/update/delete) | | `limit` | number | No | Max results (default 10, max 100) | ### salesforce\_query Execute SOQL queries against Salesforce. | Parameter | Type | Required | Description | | --------- | ------ | -------- | ----------------- | | `soql` | string | Yes | SOQL query string | ### salesforce\_crm Insert, update, or describe Salesforce objects. | Parameter | Type | Required | Description | | ------------- | ------ | -------- | -------------------------------- | | `action` | enum | Yes | `insert`, `update`, `describe` | | `object_type` | string | Yes | Salesforce object API name | | `id` | string | No | Record ID (for update) | | `data` | object | No | Field values (for insert/update) | ### ga4\_report Run Google Analytics 4 reports. | Parameter | Type | Required | Description | | ------------- | --------- | -------- | ---------------------- | | `property_id` | string | Yes | GA4 property ID | | `date_range` | object | Yes | `{startDate, endDate}` | | `metrics` | string\[] | Yes | Metric names | | `dimensions` | string\[] | No | Dimension names | | `limit` | number | No | Max rows (default 100) | ### gsc\_performance Query Google Search Console performance data. | Parameter | Type | Required | Description | | ------------ | ------- | -------- | -------------------------------------------- | | `site_url` | string | Yes | Search Console property URL | | `start_date` | string | Yes | YYYY-MM-DD | | `end_date` | string | Yes | YYYY-MM-DD | | `dimensions` | enum\[] | No | `query`, `page`, `country`, `device`, `date` | ### semrush\_research SEMrush domain analytics and keyword research. | Parameter | Type | Required | Description | | ---------- | ------ | -------- | ------------------------ | | `type` | enum | Yes | Report type | | `domain` | string | Yes | Domain to analyze | | `database` | string | No | Country (default `us`) | | `limit` | number | No | Max results (default 10) | ### gmail Search, read, or send Gmail messages. | Parameter | Type | Required | Description | | ------------ | ------ | -------- | ------------------------- | | `action` | enum | Yes | `search`, `read`, `send` | | `query` | string | No | Search query (for search) | | `message_id` | string | No | Message ID (for read) | | `to` | string | No | Recipient (for send) | | `subject` | string | No | Subject (for send) | | `body_text` | string | No | Body (for send) | ### google\_calendar Google Calendar event CRUD. | Parameter | Type | Required | Description | | ----------------------- | ------ | -------- | ------------------------------------------- | | `action` | enum | Yes | `list`, `get`, `create`, `update`, `delete` | | `calendar_id` | string | No | Default `primary` | | `time_min` / `time_max` | string | No | ISO timestamps (for list) | | `summary` | string | No | Event title (for create/update) | ### perplexity\_search Web-grounded search via Perplexity Sonar models. | Parameter | Type | Required | Description | | --------- | ------ | -------- | ------------------------------------------------------------------ | | `query` | string | Yes | Search query | | `model` | enum | No | `sonar`, `sonar-pro`, `sonar-reasoning-pro`, `sonar-deep-research` | ### aws\_bedrock\_invoke Invoke AWS Bedrock models (raw API). | Parameter | Type | Required | Description | | ------------ | ------ | -------- | ------------------------------------------------------------------- | | `model_id` | string | Yes | Bedrock model ID (e.g. `anthropic.claude-3-5-sonnet-20241022-v2:0`) | | `prompt` | string | Yes | Input prompt | | `max_tokens` | number | No | Max output tokens (default 1024) | ### aws\_bedrock\_converse Bedrock Converse API with tool use support. | Parameter | Type | Required | Description | | ---------- | ------ | -------- | -------------------- | | `model_id` | string | Yes | Bedrock model ID | | `messages` | array | Yes | Conversation history | | `system` | string | No | System prompt | | `tools` | array | No | Tool definitions | ### aws\_textract\_analyze Extract text and tables from PDFs and images. | Parameter | Type | Required | Description | | --------------- | ------- | -------- | ------------------------------ | | `document_url` | string | Yes | URL of the document to analyze | | `feature_types` | enum\[] | No | `TABLES`, `FORMS`, `QUERIES` | ### aws\_ses\_send Send transactional emails via AWS SES. | Parameter | Type | Required | Description | | ----------- | --------- | -------- | ---------------------------------- | | `to` | string\[] | Yes | Recipient addresses | | `subject` | string | Yes | Email subject | | `html_body` | string | No | HTML content | | `text_body` | string | No | Plain text content | | `from` | string | No | Sender (defaults to tenant config) | ### aws\_nova\_canvas Generate images via AWS Nova Canvas. | Parameter | Type | Required | Description | | ----------------- | ------ | -------- | --------------------------- | | `prompt` | string | Yes | Image generation prompt | | `negative_prompt` | string | No | What to exclude | | `width` | number | No | Image width (default 1024) | | `height` | number | No | Image height (default 1024) | ### gamma\_generate Generate presentations and documents via Gamma. | Parameter | Type | Required | Description | | --------- | ------ | -------- | ------------------------------------- | | `title` | string | Yes | Document title | | `content` | string | Yes | Content outline or brief | | `type` | enum | No | `presentation`, `document`, `webpage` | ### linkedin\_ads LinkedIn campaign and ad management. | Parameter | Type | Required | Description | | ------------ | ------ | -------- | -------------------------------------------------- | | `action` | enum | Yes | `list_campaigns`, `get_campaign`, `list_creatives` | | `account_id` | string | Yes | LinkedIn ad account ID | ### meta\_ads Meta (Facebook/Instagram) Ads API. | Parameter | Type | Required | Description | | ------------ | ------ | -------- | ------------------------------------------------ | | `action` | enum | Yes | `list_campaigns`, `get_insights`, `list_ad_sets` | | `account_id` | string | Yes | Meta ad account ID | | `date_range` | object | No | `{since, until}` YYYY-MM-DD | ### microsoft\_ads Microsoft Advertising (Bing Ads) API. | Parameter | Type | Required | Description | | ------------- | ------ | -------- | ----------------------------------------------- | | `action` | enum | Yes | `list_campaigns`, `get_report`, `list_accounts` | | `customer_id` | string | No | Microsoft Ads customer ID | ### microsoft\_calendar Microsoft 365 Calendar operations. | Parameter | Type | Required | Description | | ----------------------- | ------ | -------- | ------------------------------------------- | | `action` | enum | Yes | `list`, `get`, `create`, `update`, `delete` | | `calendar_id` | string | No | Default `primary` | | `time_min` / `time_max` | string | No | ISO timestamps (for list) | ### dealcloud DealCloud CRM operations (PE/VC). | Parameter | Type | Required | Description | | ------------- | ------ | -------- | ---------------------------------- | | `action` | enum | Yes | `query`, `get`, `create`, `update` | | `entity_type` | string | Yes | DealCloud entity type | | `filters` | object | No | Query filters | ### submit\_feedback Submit quality feedback on a tool call output. | Parameter | Type | Required | Description | | ----------- | ------ | -------- | ---------------------------- | | `tool_name` | string | Yes | Name of the tool being rated | | `rating` | enum | Yes | `good`, `bad`, `neutral` | | `rationale` | string | Yes | Reason for the rating | | `call_id` | string | No | Tool call ID for correlation | ### web\_fetch Fetch and parse a web page. | Parameter | Type | Required | Description | | --------- | ------ | -------- | ------------------------------------ | | `url` | string | Yes | URL to fetch | | `format` | enum | No | `markdown` (default), `html`, `text` | ### search\_knowledge Search the tenant knowledge base (Vectorize index). | Parameter | Type | Required | Description | | ----------- | ------ | -------- | ----------------------------- | | `query` | string | Yes | Search query | | `top_k` | number | No | Results to return (default 5) | | `namespace` | string | No | Knowledge namespace | ## Platform Tools (10) ### agent\_state Store and retrieve key-value state scoped to the current tenant. | Parameter | Type | Required | Description | | ----------- | ------ | -------- | ------------------------------------- | | `action` | enum | Yes | `store`, `retrieve`, `delete`, `list` | | `key` | string | No | State key | | `value` | any | No | Value to store | | `ttl_hours` | number | No | Auto-expire TTL (default 24h) | ### batch\_execute Execute up to 10 API calls in parallel. | Parameter | Type | Required | Description | | ---------- | ------- | -------- | ---------------------------------- | | `requests` | array | Yes | Array of request objects (max 10) | | `parallel` | boolean | No | Execute in parallel (default true) | ### discover\_apis Returns all connected providers with example endpoints. No parameters required. ### call\_api Generic authenticated proxy for any registered domain. | Parameter | Type | Required | Description | | --------- | ------ | -------- | ------------------------- | | `domain` | string | Yes | API domain name | | `method` | enum | No | HTTP method (default GET) | | `path` | string | Yes | API path | | `body` | any | No | Request body | | `query` | object | No | Query parameters | ### context\_query Query the context plane for facts, signals, and entity data. | Parameter | Type | Required | Description | | ------------- | ------ | -------- | ------------------------ | | `query` | string | Yes | Natural language query | | `entity_type` | string | No | Filter by entity type | | `limit` | number | No | Max results (default 10) | ### context\_explain Explain context plane reasoning and signal evaluations. | Parameter | Type | Required | Description | | ----------- | ------ | -------- | -------------------------- | | `signal_id` | string | No | Specific signal to explain | | `entity_id` | string | No | Entity to explain | ### ai\_invoke Route LLM calls through the AI Gateway with automatic model selection, observability, and budget cap. Provider and model are chosen automatically based on `intent` and `quality` — do not specify them directly. | Parameter | Type | Required | Description | | ---------------- | ------- | -------- | ------------------------------------------------------ | | `messages` | array | Yes | Conversation messages (`{role, content}` objects) | | `system` | string | No | System prompt | | `intent` | enum | No | `auto` (default), `search`, `document`, `code`, `bulk` | | `quality` | enum | No | `economy`, `standard` (default), `premium` | | `needs_web` | boolean | No | Enable web-grounded retrieval (default `false`) | | `context_tokens` | number | No | Estimated input context size for routing | | `max_tokens` | number | No | Max output tokens (default 1024, max 65536) | | `json_mode` | boolean | No | Force JSON output (default `false`) | | `temperature` | number | No | Sampling temperature (0–2) | ### llm\_invoke Direct LLM invocation with explicit provider and model selection, or shortcut tier routing. | Parameter | Type | Required | Description | | ----------------- | --------- | -------- | ------------------------------------------------------------------------------------------------------------------ | | `messages` | array | Yes | Conversation messages (`{role, content}` objects) | | `tier` | enum | No | `bulk` (Workers AI / Qwen3-30B), `standard` (DeepSeek V4 Flash), `frontier` (requires explicit `provider`+`model`) | | `provider` | enum | No | `workers_ai`, `deepseek`, `openrouter`, `groq`, `cerebras`, `gemini` | | `model` | string | No | Model ID (required when `tier: frontier`) | | `max_tokens` | number | No | Max output tokens | | `stop` | string\[] | No | Stop sequences | | `response_format` | object | No | `{type: "text" \| "json_object"}` | ### gemini\_invoke Invoke Google Gemini models directly. | Parameter | Type | Required | Description | | ------------ | ------ | -------- | ------------------------------------ | | `model` | string | Yes | Gemini model (e.g. `gemini-2.5-pro`) | | `prompt` | string | Yes | Input prompt | | `max_tokens` | number | No | Max output tokens | ### claude Invoke Claude models for nested agent calls, batch processing, file operations, or sub-agent spawning. Uses a discriminated union on `action`. **`action: "invoke"` (most common)** | Parameter | Type | Required | Description | | ------------ | ------ | -------- | ------------------------------------------------------ | | `action` | string | Yes | Must be `"invoke"` | | `messages` | array | Yes | Conversation messages (`{role, content}` objects) | | `model` | string | No | Claude model ID (default `claude-sonnet-4-6-20250514`) | | `system` | string | No | System prompt | | `max_tokens` | number | No | Max output tokens | | `tools` | array | No | Tool definitions for tool use | Other `action` values: `"batch"`, `"files"`, `"agents"` — each has its own parameter set. ## Common Parameters All curated tools support: | Parameter | Type | Default | Description | | ------------ | --------- | -------------- | --------------------------------------------- | | `fields` | string\[] | — | Dot-notation field paths to return | | `format` | enum | `full` | `full` or `compact` (strips metadata wrapper) | | `account_id` | string | tenant default | Multi-account selector | | `rationale` | string | — | Quality harness: reason for calling this tool |