Talkwalker MCP server for your AI

Change one document in the project, found by url.

Create the document if its url is new, otherwise update it.

Count results of a query in the global index, over time or over a metric.

Write an AI summary of the first 1000 results of a query in the global index, sorted by engagement, without returning the results.

Count results in the connected project, over time or over a metric, for the whole project or for chosen topics.

Write an AI summary of the first 1000 results in the connected project, sorted by engagement, without returning the results.

Rank the authors or sources with the most results in the connected project, with reach, engagement and sentiment counts.

List the Talkwalker projects this access token can reach, with their ids and account names.

See what this Talkwalker connection can query in one call: the project, its topics, filters, channels, panels and datasets with their ids, and the credits left this billing period.

Search Talkwalker's global index, outside any project (Quick Search): up to 500 sorted results from the last 30 days.

Search documents in the connected Talkwalker project, with its full history: up to 10,000 sorted results.

Get an one-call picture of a brand or topic over a period: daily volume by sentiment, overall sentiment, top countries, top themes and, in a project, top authors.

Create a stream, or replace the rules of an existing one.

Copy past results of the connected project into a collector, spending 1 credit per result, also on repeat exports.

Copy past results into a collector so they can be read in pages: makes an empty collector when none is named, then starts the export.

Check one export task: status (queued, running, finished, failed, deleted, aborted, result_limit_reached), progress from 0 to 1, processed and ready counts.

List recent export tasks with status, progress and target collector.

List the streams and collectors of the API subscription, with their state.

Listen to the connected project for up to wait_seconds and return what arrives, up to max_hits (default 20, at most 200).

Listen to a stream for up to wait_seconds and return what arrives, up to max_hits (default 20, at most 200).

See how an export is doing and read its results: status and progress of the task, then up to limit results from its collector, free once exported.

Create a collector, or change an existing one's query (its stored results stay).

Read a collector's definition and state: ACTIVE, PAUSED, ERROR, DELETED, NO_CREDITS or UNKNOWN.

Pause a collector so it stops gathering live results, and stops spending credits on them.

Read results stored in a collector, for free once collected, from resume_offset (default earliest) up to limit.

Resume a paused collector.

Read the definition of one topic or custom filter.

Create or edit topics and custom filters: this changes what the project monitors, and overwrites edits made in Talkwalker.

List every topic and custom filter of the project with its full definition, including queries.

List the custom tags defined in the project, with their ids.

List the project's topics, custom filters, channels, events, panels and datasets with their ids: the ids that search, histogram and stream tools take.

List the project's panel groups and panels with their members.

List the dashboards, reports and alerts defined in the project, with their ids and last update times.

Find logos, objects or a scene in an image at a URL.

See credits used and left this billing period, the next reset (next_billing_period, epoch ms) and six months of history.

Add one document to the project.

Add one review to a dataset of a Customer Intelligence project.

Remove one document from the project, by url.

Remove one review from a dataset, by url.

Create, update or delete several documents in the project at once.

Create, update or delete several reviews in a dataset of a Customer Intelligence project.

Bring back a document removed with talkwalker_delete_document, by url.

Bring back a review removed with talkwalker_delete_review, by url.

Change one review in a dataset, found by url.

Create the review if its url is new, otherwise update it.

Add custom metrics to a project.

Add custom fields to a dataset template in a Customer Intelligence project.

Delete one custom metric by name.

Delete several custom metrics by name.

Delete one custom field of a dataset template by name.

Delete several custom fields of a dataset template by name.

List the custom metrics of a project.

List the custom fields of a dataset template in a Customer Intelligence project.

Change the title, description or visibility of custom metrics, found by name.

Stop a running export task.

Copy past results of a query into a collector, spending 1 credit per result, also on repeat exports.

Copy past results matching a stream's rules into a collector, spending 1 credit per result, also on repeat exports.

Delete a stream.

Read a stream's rules and status.

The same as talkwalker_listen_project with the query in body.q, for a query too long for a URL.

The same as talkwalker_listen_stream with the query in body.q, for a query too long for a URL.

Delete a collector and everything stored in it.

Delete an empty topic or filter category.

Delete topics or custom filters.

Create empty panel groups.

Create source or author panels, used to filter results.

Find logos, objects or a scene in an image you provide.