Track your brand in AI answers: ChatGPT, Gemini, Claude, AI Overviews. Also a ChatGPT plugin.
Scored across 65 tools
Most tools have clearly distinct resource/action targets, and the descriptions carefully differentiate similar analytics endpoints by scope and metric. The main ambiguity is among the many analytics tools (e.g., list-brand-analytics vs. the various get-*-time-series tools, and source/source-URL breakdowns), where an agent must parse scope carefully to choose correctly.
Tool names consistently use kebab-case verb-noun phrasing, with predictable verbs such as create, get, list, update, delete, attach, and detach. Compound names like attach-brand-to-segment-tool and get-share-of-voice-time-series-tool remain readable and follow the same convention.
Sixty-five tools is too many for a single MCP server surface, even for a broad analytics platform. While each tool appears individually scoped, the large number of analytics variants creates a heavy surface that could be consolidated or grouped to improve agent usability.
The server covers core lifecycle operations for organizations, projects, monitors, prompts, segments, tags, users, and invitations, plus extensive analytics and citation data. Minor gaps exist, such as no dedicated brand deletion tool and no obvious way to list pending invitations, though these may be intentional workarounds via ignore/update flows.
Attach a brand to a segment. Idempotent: attaching a brand that is already attached succeeds without change. Both the brand and the segment must belong to the same project. Membership is what makes a brand a competitor, so attaching a discovered brand starts tracking it and attaching an ignored brand revives it, clearing its ignored mark. The own brand is implicit in every segment, so attaching it changes nothing.
Inputs: brand_id (required), segment_id (required).
Attach a tag to a prompt. Idempotent: attaching a tag that is already attached succeeds without change. Both the prompt and the tag must belong to the same project.
Inputs: tag_id (required), prompt_id (required).
Create a brand in a project, ahead of Pineprompt discovering it from an AI response. Segment membership is what makes a brand a competitor, so pass segment_ids to start tracking it right away; without them the brand starts out discovered. When aliases are omitted, the brand is seeded with its display name. Display names are unique per project, case-insensitive, so a name that collides with an existing brand's display name or alias returns an error. To designate a brand as the project's own brand, pass its id as brand_id on the update-project tool instead.
Inputs: project_id (required), display_name (required), aliases, domains, +1 more.
Invite a person to join an organization by email. This sends an invitation email to the given address, outside Pineprompt. Requires admin or owner rights in the organization. The recipient may accept it within 14 days. Allowed roles are admin and member; owner cannot be invited. Pending invitations count toward the organization's plan seat quota.
Inputs: role (required), email (required), organization_id (required).
Active prompts are sent daily to third-party AI search engines (for example ChatGPT, Gemini and Perplexity) on Pineprompt's schedule, which uses plan quota. Create a new monitor for a project. Each monitor runs prompts against one or more AI platforms in the given country and language. Subject to the organization's plan quota.
Inputs: name (required), country (required), language (required), is_active (required), +2 more.
Create a new organization with the authenticated user as the owner. The returned organization id may be used as organization_id in subsequent tool calls.
Inputs: name (required).
Create a new project in the given organization. Subject to the organization's plan quota.
Inputs: name (required), organization_id (required).
Active prompts are sent daily to third-party AI search engines (for example ChatGPT, Gemini and Perplexity) on Pineprompt's schedule, which uses plan quota. Create a new prompt for a monitor. The prompt will run against the monitor's configured AI platforms on schedule when is_active is true. Subject to the organization's plan quota.
Inputs: prompt (required), is_active (required), monitor_id (required).
Create a new segment for a project. Segment names are unique per project, case-insensitive, so sending a duplicate (in any casing) returns an error.
Inputs: name (required), project_id (required).
Create a new tag for a project. Requires a name (unique per project, case-insensitive, so sending a duplicate in any casing returns an error) and a color from the preset palette.
Inputs: name (required), color (required), project_id (required).
Cancel a pending invitation.
Inputs: invitation_id (required).
Delete a monitor. The monitor stops running immediately. The monitor and its historical responses are soft-deleted, retained for 30 days, and then permanently removed.
Inputs: monitor_id (required).
Delete an organization. Its subscription is canceled immediately. The organization and everything in it, including projects, monitors, and historical data, is soft-deleted, retained for 30 days, and then permanently removed.
Inputs: organization_id (required).
Delete a project. Its monitors stop running immediately. The project and everything beneath it, including its historical data, is soft-deleted, retained for 30 days, and then permanently removed.
Inputs: project_id (required).
Delete a prompt. The prompt stops running immediately. The prompt and its historical responses are soft-deleted, retained for 30 days, and then permanently removed.
Inputs: prompt_id (required).
Delete a segment. Attached brands are untouched but the segment stops appearing on them; a brand whose only segment this was derives back to discovered. The segment is soft-deleted, retained for 30 days, and then permanently removed. A project must keep at least one segment, so deleting the last one returns an error.
Inputs: segment_id (required).
Delete a tag. Attached prompts are untouched but the tag stops appearing on them. The tag is soft-deleted, retained for 30 days, and then permanently removed.
Inputs: tag_id (required).
Remove a user from an organization. The organization owner cannot be removed.
Inputs: user_id (required), organization_id (required).
Detach a brand from a segment. Idempotent: detaching a brand that is not attached succeeds without change. Both the brand and the segment must belong to the same project. Membership is what makes a brand a competitor, so detaching one from its only segment stops tracking it and it derives back to discovered, returning to the discover pool with its mention history intact. The own brand is implicit in every segment and cannot be detached.
Inputs: brand_id (required), segment_id (required).
Detach a tag from a prompt. Idempotent: detaching a tag that is not attached succeeds without change. Both the prompt and the tag must belong to the same project.
Inputs: tag_id (required), prompt_id (required).
Fetch a single brand mention by its numeric id. If you would like to embed the brand it points to, or a deeper parent chain, you may pass include.
Inputs: brand_mention_id (required), include.
Fetch a single citation by its numeric id. Optionally pass include to load parent chains.
Inputs: citation_id (required), include.
Fetch a single monitor by its numeric id. Optionally pass include to load the parent project chain (project, project.organization).
Inputs: monitor_id (required), include.
Fetch a single organization by its numeric id.
Inputs: organization_id (required).
Average position of each tracked brand within the answers that mention it, per bucket. Position counts brands in the order they appear in an answer, so a lower number is more prominent and 1 is the first brand named. Scope by brand_id for one line; by segments for that segment's brands; otherwise top top_n by mention count (default 20, max 50). Per-bucket value is null when the brand had no mentions in that bucket.
Inputs: end_date (required), project_id (required), start_date (required), tags, +7 more.
Fetch a single project by its numeric id. Optionally pass include to load the parent organization.
Inputs: project_id (required), include.
Fetch a single prompt by its numeric id. Optionally pass include to load the parent monitor/project/organization chain.
Inputs: prompt_id (required), include.
Fetch a single AI response by its numeric id. Optionally pass include to load parent chains. Brand mentions and citations are not embedded; query /v1/brand-mentions and /v1/citations separately.
Inputs: response_id (required), include.
Fetch a single segment by its numeric id. Optionally pass include to load the parent project chain.
Inputs: segment_id (required), include.
Average sentiment score (0-100) of each tracked brand's mentions per bucket. Scope by brand_id for one line; by segments for that segment's brands; otherwise top top_n by mention count (default 20, max 50). Per-bucket value is null when the brand had no mentions in that bucket.
Inputs: end_date (required), project_id (required), start_date (required), tags, +7 more.
Fetch a single source by its numeric id. Optionally pass include to load the parent project chain.
Inputs: source_id (required), include.
Citation breakdown by source-type classification (e.g. blog, news, forum). Unpaginated: one row per source-type enum value. Cannot be scoped by segment.
Inputs: end_date (required), project_id (required), start_date (required), tags, +4 more.
Citation breakdown by URL-type classification (e.g. article, documentation, product_page). Unpaginated: one row per URL-type enum value. Cannot be scoped by segment.
Inputs: end_date (required), project_id (required), start_date (required), tags, +4 more.
For each top URL, the percentage of analyzed responses in the bucket that cited that exact URL. Capped to the top top_n URLs (default 20, max 50).
Inputs: end_date (required), project_id (required), start_date (required), tags, +5 more.
For each top source, the percentage of analyzed responses in the bucket that cited any URL from that source. Capped to the top top_n sources (default 20, max 50).
Inputs: end_date (required), project_id (required), start_date (required), tags, +5 more.
Fetch a single tag by its numeric id. Optionally pass include to load the parent project chain.
Inputs: tag_id (required), include.
Returns the share of analyzed responses that mention each tracked brand, bucketed by interval (day/week/month/year). Scope by brand_id for one line; by segments for that segment's brands; otherwise top top_n by mention count (default 20, max 50).
Inputs: end_date (required), project_id (required), start_date (required), tags, +7 more.
Per-brand aggregates over the filter window: visibility, share-of-voice, average sentiment, and average position. Returns one row for every tracked brand in scope; brands with no mentions in the window appear with zero visibility and share-of-voice and a null sentiment and position. Paginated.
Inputs: end_date (required), project_id (required), start_date (required), page, +9 more.
List the brand mentions detected across a project, newest first. Returns up to 50 per page; follow nextCursor to fetch additional pages. Optionally pass include to load parent chains (response, brand, project, organization).
Inputs: project_id (required), cursor, include, brand_id, +1 more.
List the brands configured for a project, newest first. Returns up to 50 per page; follow nextCursor to fetch additional pages. Optionally pass include to load the parent project chain.
Inputs: project_id (required), cursor, include.
List the citations collected across a project, newest first. Each citation carries the URL the platform linked plus its domain. Returns up to 50 per page; follow nextCursor to fetch additional pages. Optionally pass include to load parent chains (response, source_url, source, project).
Inputs: project_id (required), cursor, include, response_id.
Per-monitor aggregates (visibility, share-of-voice, sentiment, position) scoped to a single brand. Paginated.
Inputs: brand_id (required), end_date (required), project_id (required), start_date (required), +10 more.
List the monitors configured for a project, newest first. Returns up to 50 per page; follow nextCursor to fetch additional pages. Optionally pass include to load the parent project chain (project, project.organization).
Inputs: project_id (required), cursor, include.
List the organizations you belong to, newest first. Returns up to 50 per page; follow nextCursor to fetch additional pages. Each item includes the organization's id, name, and the user's role. Use this as the entry point to discover ids for other tools.
Inputs: cursor.
Per-AI-platform aggregates (visibility, share-of-voice, sentiment, position, used_count) scoped to a single brand. One row per platform present in the filtered response set; sorted by visibility desc.
Inputs: brand_id (required), end_date (required), project_id (required), start_date (required), +6 more.
List the projects within an organization, newest first. Returns up to 50 per page; follow nextCursor to fetch additional pages. Optionally pass include to load the parent organization.
Inputs: organization_id (required), cursor, include.
Per-prompt aggregates (visibility, share-of-voice, sentiment, position) scoped to a single brand, plus the distinct tracked brands mentioned in each prompt's responses during the window. Paginated. To fetch more rows, request the next page; the envelope reports the total under meta.last_page.
Inputs: brand_id (required), end_date (required), project_id (required), start_date (required), +10 more.
List the prompts configured for a monitor, newest first. Returns up to 50 per page; follow nextCursor to fetch additional pages. Optionally pass include to load the parent monitor/project/organization chain.
Inputs: monitor_id (required), cursor, include.
List the AI responses collected for a project, newest first. Returns up to 50 per page; follow nextCursor to fetch additional pages. Optionally pass prompt_id to return only the responses for a single prompt, or source_url_id to return only the responses that cite a given source URL. Pass include to load parent chains (prompt, monitor, project, organization). Brand mentions and citations are not embedded; query /v1/brand-mentions and /v1/citations separately.
Inputs: project_id (required), cursor, include, prompt_id, +1 more.
List the segments configured for a project, newest first. Returns up to 50 per page; follow nextCursor to fetch additional pages. Optionally pass include to load the parent project chain.
Inputs: project_id (required), cursor, include.
Per-source aggregates: the percentage of analyzed responses in the window that cited each domain, and the average number of citations from each domain per analyzed response. Paginated.
Inputs: end_date (required), project_id (required), start_date (required), page, +8 more.
List the sources cited by AI platforms when responding to a project's prompts, ordered by domain. Returns up to 50 per page; follow nextCursor to fetch additional pages. Optionally pass include to load the parent project chain.
Inputs: project_id (required), cursor, include.
Per-URL aggregates: the percentage of analyzed responses in the window that cited each URL, and the average citation count for each URL per analyzed response. Paginated.
Inputs: end_date (required), project_id (required), start_date (required), page, +8 more.
Per-prompt citation analytics scoped to one source URL: how many distinct analyzed responses on each prompt cited the URL in the window, plus the most recent citation date. Paginated.
Inputs: end_date (required), project_id (required), start_date (required), source_url_id (required), +9 more.
List the tags configured for a project, newest first. Returns up to 50 per page; follow nextCursor to fetch additional pages. Optionally pass include to load the parent project chain.
Inputs: project_id (required), cursor, include.
List the members of an organization, ordered by name. Returns up to 50 per page; follow nextCursor to fetch additional pages.
Inputs: organization_id (required), cursor.
Overwrites the given fields; previous values are not kept. Update a brand's display name, aliases, domains, or ignored state. Every field is optional, so send only the fields to change; the aliases and domains lists replace the current ones in full. A brand's kind is derived from context rather than written: pass ignored: true to stop tracking a brand (it is detached from every segment as the mark is set), ignored: false to return an ignored brand to the discovered pool, attach it to a segment to make it a competitor, or designate it through the update-project tool's brand_id to make it the own brand. The own brand cannot be ignored.
Inputs: brand_id (required), aliases, domains, ignored, +1 more.
Active prompts are sent daily to third-party AI search engines (for example ChatGPT, Gemini and Perplexity) on Pineprompt's schedule, which uses plan quota. Overwrites the given fields; previous values are not kept. Update an existing monitor. Every field is optional, so send only the fields to change; a field that is omitted keeps its current value. Platforms sent without a country are validated against the monitor's stored country, and a country sent without platforms must be supported by every platform the monitor already runs on.
Inputs: monitor_id (required), name, country, language, +2 more.
Overwrites the given fields; previous values are not kept. Update an existing organization. Every field is optional; a field that is omitted keeps its current value.
Inputs: organization_id (required), name.
Overwrites the given fields; previous values are not kept. Update an existing project. Every field is optional, so send only the fields to change; a field that is omitted keeps its current value. Pass brand_id to designate one of the project's brands as its own brand, the one visibility is anchored against. The own brand is implicit in every segment and is never ignored, so it leaves any segments and sheds any ignored mark as it takes the role; a previous own brand derives back to discovered. Pass null to leave the project without an own brand.
Inputs: project_id (required), name, brand_id.
Active prompts are sent daily to third-party AI search engines (for example ChatGPT, Gemini and Perplexity) on Pineprompt's schedule, which uses plan quota. Overwrites the given fields; previous values are not kept. Update an existing prompt. Every field is optional, so send only the fields to change; a field that is omitted keeps its current value. Pass monitor_id to move the prompt to another monitor in the same project (collected responses move with it), or tags to replace the prompt's full set of tag ids.
Inputs: prompt_id (required), tags, prompt, is_active, +1 more.
Overwrites the given fields; previous values are not kept. Rename an existing segment. The name is optional, so a call that omits it leaves the segment unchanged. Names are unique per project, case-insensitive, so renaming to a duplicate (in any casing) returns an error.
Inputs: segment_id (required), name.
Overwrites the given fields; previous values are not kept. Update an existing tag's name and color. Every field is optional, so send only the fields to change; a field that is omitted keeps its current value. Names are unique per project, case-insensitive, so renaming to a duplicate (in any casing) returns an error.
Inputs: tag_id (required), name, color.
Overwrites the given fields; previous values are not kept. Change a user's role within an organization. Allowed roles are admin and member; the organization owner's role cannot be changed and users cannot be promoted to owner.
Inputs: role (required), user_id (required), organization_id (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Track brand visibility in ChatGPT, Gemini and Google AI answers: competitors, cited sources, checks.
Track and manage your brand's visibility in AI answers: ChatGPT, Gemini, Google AI, Copilot.
Track your brand's visibility in ChatGPT, Perplexity, Claude, Gemini, Grok and DeepSeek answers.
Track how ChatGPT, Claude, Perplexity and other AI assistants mention your brand, right from your AI client.
Add one secure layer between your agents and this server.
