MCP App Store
Productivity
Gong icon

Gong

by Lior Ben Ami

Overview

Gong brings deep revenue insights into your Copilot workflows, helping teams ask questions, generate briefs, and surface risks, objections, next steps, and stakeholder context from Gong AI OS data.

Tools

ask_account

ChatGPT
Tool Name: ask_account Description: The ask_account tool answers natural-language questions about a specific CRM account by analyzing Gong activities (calls and messages) within a defined time range. It is designed to provide synthesized insights, not raw data. You provide the CRM account (by name or ID), a free-text question, and optionally a time window. The tool returns an answer derived from the relevant account activities during that period. If no date range is provided, the tool defaults to the last 30 days. To specify a custom date range, provide both fromDateTime and toDateTime in ISO 8601 format with timezone (e.g. 2025-08-01T00:00:00Z). The workspace field is optional — if your account has only one Gong workspace it will be selected automatically. If you have multiple workspaces, provide either a numeric workspace ID or an exact workspace name. When to use: - When you need summaries, insights, or themes about a single CRM account. - To understand account progress, deal blockers, objections, risks, stakeholders, or next steps. - To extract high-level context from calls and messages linked to one CRM account. When NOT to use: - Do NOT use to fetch raw data such as transcripts, message bodies, or activity lists. - Do NOT use for analytics across multiple accounts or workspace-level reporting. - Do NOT use to modify account data or trigger workflows. Key Characteristics: - Read-only: The tool never modifies data. - Time-bounded: Considers only calls and messages within the date range (defaults to last 30 days if not specified). - Summarization-focused: Produces synthesized insights, not verbatim extracts. - Single-shot: Does not support pagination or streaming. Usage Tips: - If your account has two or more workspaces, providing the workspace parameter is mandatory; either provide a numeric workspace ID (e.g. "1234567890") or an exact workspace name (case-insensitive, e.g. "Sales"). If the value can be parsed as a number it is used as a workspace ID; otherwise it is resolved as a workspace name. - crmAccount accepts either the full account name (e.g. 'Cisco Systems Inc') or its direct CRM account ID. Name lookup uses exact matching — provide the full name as it appears in the CRM. - Date range is optional. If not provided, defaults to the last 30 days. To specify a custom range, provide both fromDateTime and toDateTime in ISO 8601 format with timezone (e.g. "2025-08-01T00:00:00Z"). Both must be provided together — providing only one will result in an error. - Ask concise, insight-oriented questions instead of requesting raw data. - The tool works best when targeting a single account and a specific analysis goal. Clarification Flow: When multiple CRM accounts match the given name, the tool returns an error result containing a JSON list of matches sorted by most recent activity. Each match has three fields: - crmId: use this value as crmAccount when re-calling the tool. - name: the account name in CRM. - lastActivity: ISO 8601 timestamp of the most recent Gong activity, or null if unknown. Present the matches to the user — highlight accounts with a recent lastActivity as active — ask them to pick one, and re-call with the chosen crmId. Not Found Flow: When no CRM account matches the given name or ID, the tool returns an error with reason=CRM_ENTITY_NOT_FOUND. Inform the user that the account was not found and ask them to verify the name or ID as it appears in their CRM, then re-call with the corrected value. Example Questions: - "What are the top objections raised by this account in the past 30 days?" - "Who are the key stakeholders and what are their goals?" - "What blockers are delaying the deal close?" - "Summarize the main themes from calls in August." Example requests: ----------------- Example 1 — Using default date range (last 30 days) ---------------------------------------------------- { "crmAccount": "0015g00000ABCDeQAL", "question": "What are the main objections raised by the prospect recently?" } Example 2 — Custom date range ------------------------------- { "workspace": "1234567890", "crmAccount": "0…

ask_deal

ChatGPT
Tool Name: ask_deal Description: The ask_deal tool answers natural-language questions about a specific CRM deal or opportunity by analyzing Gong activities (calls and messages) within a defined time range. It is designed to provide synthesized insights, not raw data. You provide the CRM deal by name (e.g. "Acme Corp - Enterprise Q4") or direct CRM ID (e.g. "006EXAMPLE00001"), a free-text question, and optionally a time window. The tool returns synthesized insights derived from the relevant activities during that period. If no date range is provided, the tool defaults to the last 30 days. To specify a custom date range, provide both fromDateTime and toDateTime in ISO 8601 format with timezone (e.g. 2025-08-01T00:00:00Z). The workspace field is optional. If the account has exactly one Gong workspace it is selected automatically. If multiple workspaces exist, provide either a numeric workspace ID or an exact workspace name (case-insensitive). When to use: - When you need insights, themes, or a summary about a single CRM deal or opportunity. - To understand deal momentum, buyer objections, competitive mentions, or negotiation progress. - To identify key stakeholders, their concerns, and agreed-upon next steps for a deal. - To extract high-level context from calls and messages linked to one CRM deal. When NOT to use: - Do NOT use to fetch raw data such as transcripts, message bodies, or activity lists. - Do NOT use for analytics across multiple deals, pipeline reviews, or workspace-level reporting. - Do NOT use to modify deal data or trigger CRM workflows. - Do NOT use for account-level questions that span multiple deals — use ask_account instead. Key Characteristics: - Read-only: The tool never modifies data. - Time-bounded: Considers only calls and messages within the date range (defaults to last 30 days if not specified). - Summarization-focused: Produces synthesized insights, not verbatim extracts. - Single-shot: Does not support pagination or streaming. Usage Tips: - crmDeal accepts the CRM deal/opportunity name (e.g. 'Acme Corp - Enterprise Q4') or its direct CRM ID. Name lookup uses exact matching — provide the full deal name as it appears in the CRM. - If your account has two or more workspaces, providing the workspace parameter is mandatory. Supply either a numeric workspace ID (e.g. "1234567890") or a workspace name (case-insensitive, e.g. "Sales"). If the value parses as a number it is used as a workspace ID; otherwise it is resolved as a workspace name. - Date range is optional. If not provided, defaults to the last 30 days. To specify a custom range, provide both fromDateTime and toDateTime in ISO 8601 format with timezone (e.g. "2025-08-01T00:00:00Z"). Both must be provided together — providing only one will result in an error. - Ask concise, insight-oriented questions rather than requesting raw data. Clarification Flow: When multiple CRM deals match the given name, the tool returns an error result containing a JSON list of matches sorted by most recent activity. Each match has three fields: - crmId: use this value as crmDeal when re-calling the tool. - name: the deal name in CRM. - lastActivity: ISO 8601 timestamp of the most recent Gong activity, or null if unknown. Present the matches to the user — highlight deals with a recent lastActivity as active — ask them to pick one, and re-call with the chosen crmId. Not Found Flow: When no CRM deal matches the given name or ID, the tool returns an error with reason=CRM_ENTITY_NOT_FOUND. Inform the user that the deal was not found and ask them to verify the name or ID as it appears in their CRM, then re-call with the corrected value. Example Questions: - "What pricing concerns has the buyer raised in the last 30 days?" - "Has the prospect mentioned any competing solutions during recent calls?" - "What are the main blockers preventing this deal from closing?" - "Summarize the negotiation progress and key commitments from September." - "Who are the decision makers involved and what are their priorities?" Example requests: ----------------- Example 1 — Using defaul…

generate_brief

ChatGPT
Tool Name: generate_brief Description: The generate_brief tool creates comprehensive summarized briefs about CRM entities (accounts, deals, or contacts) by analyzing Gong activities within a specified time period. Unlike ask_account and ask_deal which answer specific questions, this tool generates structured summaries across multiple predefined categories. It produces comprehensive insights organized into different categories such as key themes, stakeholders, risks, next steps, and other relevant aspects of the entity's activities. The workspace field is optional. If the account has exactly one Gong workspace it is selected automatically. If multiple workspaces exist, provide either a numeric workspace ID or an exact workspace name (case-insensitive). When to use: - When you need a comprehensive overview of an entity's recent activities. - To generate structured summaries for account/deal reviews or handovers. - When you want insights across multiple categories rather than answers to specific questions. - For creating executive summaries or briefing documents. When NOT to use: - When you need answers to specific, targeted questions about a single account or deal. Use ask_account for account-scoped questions and ask_deal for deal-scoped questions. - Do NOT use to fetch raw data such as transcripts, message bodies, or activity lists. - Do NOT use for analytics across multiple entities or workspace-level reporting. - Do NOT use to modify entity data or trigger workflows. Key Characteristics: - Read-only: The tool never modifies data. - Time-bounded: Considers only calls and messages within the specified date range. - Multi-category: Provides insights across multiple predefined categories. - Structured output: Returns organized summaries rather than free-form answers. Response Handling: The response is a curated Gong summary. Present all sections and items verbatim. Do not summarize or reorganize the content. Show call titles as clickable links and include a source count per section. Expand email details only when the user requests it. Format for readability. Supported Entity Types: - ACCOUNT: For CRM accounts/companies - DEAL: For CRM deals/opportunities (name or CRM ID) - CONTACT: For CRM contacts/leads (CRM ID only — name resolution not supported) Supported Period Types: - CUSTOM_RANGE: Use specific fromDateTime and toDateTime - LAST_7DAYS: Activities from the last 7 days - LAST_30DAYS: Activities from the last 30 days - LAST_90DAYS: Activities from the last 90 days Usage Tips: - crmEntity accepts the entity name or CRM ID for ACCOUNT and DEAL. Name lookup uses exact matching — provide the full name as it appears in the CRM. For CONTACT, provide the direct CRM contact ID. If multiple matches are found, a clarification response is returned. - For ACCOUNT, entity names must be provided in full — partial matches are not supported. Name lookup uses exact matching — provide the full name as it appears in the CRM. - If your account has two or more workspaces, providing the workspace parameter is mandatory. Supply either a numeric workspace ID (e.g. "1234567890") or a workspace name (case-insensitive, e.g. "Sales"). If the value parses as a number it is used as a workspace ID; otherwise it is resolved as a workspace name. - periodType is optional — defaults to LAST_90DAYS if omitted. Use CUSTOM_RANGE with specific fromDateTime/toDateTime for a custom window. - briefName must exactly match the name of a published brief in the Gong workspace. It is a lookup key, not a free-text label. If the name does not match, the tool returns an error listing the available brief names and their descriptions so you can choose the correct one. Clarification Flow: When multiple CRM entities match the given name, the tool returns an error result containing a JSON list of matches sorted by most recent activity. Each match has three fields: - crmId: use this value as crmEntity when re-calling the tool. - name: the entity name in CRM. - lastActivity: ISO 8601 timestamp of the most recent Gong activity, or null if unknown. Present the match…

App Stats

3

Tools

ChatGPT

Platforms

Works with

ChatGPT

Data refreshed daily