Business & Operations

Seismic

by Seismic Software llc

Overview

The Seismic MCP Connector for ChatGPT enables secure integration between ChatGPT and Seismic, allowing ChatGPT to generate accurate, Seismic-specific responses using approved sales and enablement content. Built on the Model Context Protocol (MCP), the connector provides authentication and scope-based access control, ensuring only authorized tools and content are available within ChatGPT. To configure access, users must install the MCP Connector for ChatGPT app from Seismic Exchange, where tenant admins define scopes and permissions that control which Seismic capabilities are exposed. Once configured: ChatGPT connects to Seismic through the MCP server Only approved tools and content are accessible Responses are grounded in trusted, compliant materials Without proper installation and scope configuration, Seismic tools will not be visible in ChatGPT.

Tools

generate_livesend_link

ChatGPT

Purpose: Generates a LiveSend link with configurable settings and provided contents, enabling controlled sharing and engagement tracking. When to use: Use this endpoint when you need to create a LiveSend link for sending content to recipients, associating it with CRM context, or enabling tracking and management via the LiveSend detail page. Side effects: Creates a new LiveSend entity and persists it for tracking, analytics, and lifecycle management. Guardrails: Only authorized users can create LiveSend links. Ensure inputs comply with content access policies and recipient permissions. Do not expose external-facing URLs if restricted by configuration.

generative_search_for_answer_or_summary

ChatGPT

Purpose: Generates a natural-language answer to a user question and returns the supporting sources used. When to use: Use this tool when the user explicitly wants an answer, explanation, summary, or recommendation and expects narrative output. Use when you need both: (a) an answer and (b) citations/sources that justify it. When NOT to use: Do not use this tool if the user only wants the raw sources, links, or documents. Do not use this tool for compliance workflows where only evidence collection is required. Guardrails: Answer must be grounded only in returned sources; do not invent citations. If filters were applied and results are empty or weak, retry without any filter before reporting no results. Only include the filter parameter when the user explicitly requests filtering (for example: "show me content from this year", "filter by region", "only EMEA documents"). Do NOT infer or add filters — especially date filters — that the user did not explicitly ask for. Never add a date filter to improve answer quality or recency unless the user specifically requested it.

generative_search_for_retrieving_contents

ChatGPT

Purpose: Retrieves relevant sources for a user query without generating an answer. It Search and return the most relevant content recommendations ranked by relevance score. Note- Prefer higher score items. When to use: Use this tool when the user asks for “sources only”, “links”, “documents”, “proof”, or “what did you find”. Use this tool as the first step when you want to inspect evidence before composing an answer yourself. When NOT to use: Do not use this tool if the user expects a direct answer and you do not plan to compose one. Do not use this tool to generate summaries or conclusions (it returns sources only). Side effects: Read-only. Guardrails: Return only published content the caller is authorized to access. Do not infer conclusions; do not fabricate missing details. If no sources are found, suggest query refinements (keywords, filters, time range). Only include the filter parameter when the user explicitly requests filtering (for example: "show me content from this year", "filter by region", "only EMEA documents"). Do NOT infer or add filters — especially date filters — that the user did not explicitly ask for. Never add a date filter to improve answer quality or recency unless the user specifically requested it.