Resources · Claude connector

Claude Unwrap connectorWhat Claude can do in your Unwrap account.

The Claude Unwrap connector exposes 17 tools. 12 read your customer feedback and Unwrap's help docs, 3 produce or send something, and 2 are documented nowhere. Here is what Claude analyzes, what it produces, and where Unwrap's docs contradict themselves.

Verified Trustpilot reviews · AI, automation & growth agency

Overview

What changes when you listen to customers

Unwrap organizes your customer feedback into views, segments and theme groups. With the connector you ask Claude your question: it picks the Unwrap view to query, looks for the topics that matter, filters by segment and can come back with statistics and a chart.

Spot the priority issues. search_actionable_groups returns the most actionable groups for a question, with filters and sorting on a numeric segment. Unwrap pitches it for top issues and prioritization.

Follow a trend. plot_chart draws charts or tables over time or by breakdown, to see how something evolves, how items rank or how shares compare.

Measure what customers actually say. Claude builds a sample of feedback with sample_feedback_entries, then analyze_entry_text runs one question across every entry and returns how often an answer comes up, with examples.

What Claude does not do: edit your source feedback or your taxonomy, since no documented tool handles that. Nothing runs on its own either, so no alert when a new theme climbs. Continuous monitoring is a job for an automation tool, a different trade; Integrations gathers that side, tool by tool, for when you need something running on a schedule.

Vocabulary

The vocabulary in one minute

Five words to read the rest without detours.

Connector
The link you set up once between Claude and an account you already have, so it can work there while it answers you.
Tool
A named action the connector opens to Claude. Claude picks the ones it needs on its own; the directory sheet lists them.
Authorization
Unwrap's own sign-in screen, where you hand Claude the access it will use. You can take it back the same way.
Approval
The confirmation Claude waits for before an action that changes an account, shown in the chat right when it matters.
MCP
The shared standard behind connectors: it is what lets an assistant like Claude talk to an outside service.
Connect

Connect the Claude Unwrap connector

  1. 01

    Find Unwrap

    In Claude's settings, open Customize, then Connectors, and look for Unwrap. On a Team or Enterprise workspace, an Owner or Primary Owner has to add it first so each member can connect.

  2. 02

    Start the connection

    Click Connect on its row, then sign in through the window Unwrap opens itself. If the connection drops, use Disconnect and reconnect, which is also what Unwrap recommends.

  3. 03

    Read the authorization screen

    Read the authorization screen before you approve. It belongs to Unwrap, not to Claude, and it decides what the access covers. Whatever you grant there can be withdrawn later.

Tools

The 17 tools, grouped by what they do

Unwrap gives Claude 17 tools: 12 that read your account, 3 that change something in it, and 2 no official source describes.

Twelve reading tools, three that produce or send something, two without a description. Tool names stay exactly as Claude shows them.

  • 12 read
  • 3 write
  • 2 not documented

What Claude reads (12)

12 tools

list_views

Lists the Unwrap views your organization has, with their ID and name. It is the first thing Claude calls, since it has to know which view to work on.

When it helps
when the organization has several views and one must be picked for the question.
Watch out
only the views your permissions open to you appear in the list.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

get_started

Hands Claude Unwrap's usage guide: the data model and the recommended workflows. Unwrap asks for it to be called first, before any search.

When it helps
at the start of an analysis, so Claude understands how your feedback is organized before it looks for anything at all.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

list_feedback_integrations

Shows which feedback sources feed a given view and how well each is covered, so you know where the analyzed opinions come from.

When it helps
before drawing a conclusion, checking which feedback sources a view covers.
Watch out
coverage is read view by view, so the view to query has to be known first, which list_views provides.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

search_segments

Searches segment fields and their values by meaning for the ones tied to a question.

When it helps
narrowing the analysis to a segment without knowing the field's exact name.
Watch out
for a value's exact spelling, list_segment_field_values takes over.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

list_segment_field_values

Lists the exact values allowed for a segment field, so filters match the spelling Unwrap expects down to the character.

When it helps
a filter returns nothing because a segment name is spelled differently in Unwrap, and you want to find out how.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

discover_filter_schema

Returns every filter field available for a view, so the valid options are known up front instead of discovered by trial and error.

When it helps
before a filtered analysis, to know which criteria the view offers.
Watch out
it describes the possible filters on the view without filtering anything itself.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

search_groups

Looks through Unwrap's taxonomy of groups by meaning to find the issues or themes tied to a query, with optional filtering on top.

When it helps
finding the themes tied to a subject, whatever words the feedback happens to use for it.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

search_actionable_groups

Returns the most actionable groups for a question, with filters and sorting on a numeric segment. Unwrap aims it at top-issue lists and prioritization.

When it helps
ranking the topics behind a question by importance, filtered on a segment you choose.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

plot_chart

Produces charts or tables over time or by breakdown, to follow a trend, compare shares between groups or build a ranking.

When it helps
following a topic over time or comparing its share with the other topics in a view.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

analyze_entry_text

Puts one instruction or question to every entry in a saved sample and returns prevalence statistics along with example results.

When it helps
measuring what share of a sample matches a given question.
Watch out
it needs a sample first, made with sample_feedback_entries.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

search_documentation

Queries Unwrap's public help center and returns a ranked list of matching article titles, each with its URL.

When it helps
you wonder how to plug a new feedback source into Unwrap and want the official article rather than a guess.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

read_documentation_article

Reads one or more articles from Unwrap's documentation and returns them as Markdown, handy for setup or how-to questions.

When it helps
for a setup or integration question answered in Unwrap's help center.
Watch out
the articles come from Unwrap's public help center, which search_documentation is there to search.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

What Claude creates or sends (3)

3 tools

Three tools that create a sample or send feedback. No source spells out their confirmation, so the general rule applies.

sample_feedback_entries

Approval: see the rule

Creates a named sample of feedback entries, either random to stay representative or targeted by meaning. Later analyses run on it.

What Claude asks for
no source describes a confirmation specific to this tool; the general approvals rule applies.
When it helps
you would rather study a representative slice of comments than the entire mass.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

classify_feedback_entries

Approval: see the rule

Sorts the entries of a saved sample into labels defined in advance and creates one sub-sample per label.

What Claude asks for
the general approvals rule covers it, as no source gives it a confirmation of its own.
When it helps
when a sample must be split across categories you set yourself.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

leave_feedback

Approval: see the rule

Sends feedback straight to the Unwrap team about the connector itself, not about your customer data.

What Claude asks for
no particular confirmation is described, so the default rule is what applies.
When it helps
a tool gave you an odd answer and you want to flag it to the publisher.

Sourcedocs.unwrap.ai · September 9, 2026 ↗

Not documented (2)

2 tools

The directory publishes these two names and says nothing about their role. This page won't guess from the names.

query_feedback_data

The directory sheet publishes this name, and no official source describes it under that name. Unwrap's article mentions a tool with a similar name, without anything confirming it is the same one.

Watch out
the general approvals rule covers it.

get_group_insights_playbook

This name only appears on the directory sheet: neither Unwrap's article nor any Claude page describes it, and this page will not guess anything at all from the label.

Watch out
whatever its real role, the default confirmation described below applies.
Approvals

What Claude asks before it acts

By default, Claude stops and asks for your go-ahead before every action it takes on an account for you. The request shows up in the chat when it applies.

Unwrap recommends a setup of its own: leave its reading tools on permanent allow and keep its write tools under approval. On a Team or Enterprise workspace, owners decide whether a member may let some actions through without a new prompt, and they can cap what the connector does for the whole organization. Claude only sees the views your permissions open to you, never the rest of the account, whoever set the connector up.

Plans

Which plans it is available on

None of the 819 sheets in the official directory shows plan availability. That answer, connector by connector, is not published anywhere.

The general rule is published: remote connectors are open to all users on Claude, Cowork, Claude Desktop and mobile, and desktop extensions install on Claude Desktop. On Team and Enterprise, an Owner or Primary Owner switches the connector on for the organization before members connect. The current state is on the connector's sheet in the official directory.

Limits

Where this connector stops

A connector is not an automation: Claude calls these tools while it answers you, and nothing fires on its own. Only Unwrap documents this connector, in its knowledge base.

Unwrap's documentation does not quite match the sheet: two tool names differ or are missing, and the article contradicts itself about being read-only, as the box below explains. The directory badge is not a security audit, and Anthropic says so on every sheet: it does not choose a publisher's tools and does not guarantee how they behave. For monitoring that runs without you on a schedule, see the automation side under Integrations.

Two official sources disagree

Are all of Unwrap's tools read-only?

What this page followsUnwrap's page contradicts itself on the same date. This page follows the tool-by-tool descriptions: those three tools sit on the write side, and Unwrap advises keeping its write tools under approval anyway.

Need help

Need help connecting Unwrap to Claude?

A person reads every message.

FAQ

Claude Unwrap connector: common questions

01What can Claude do with the Claude Unwrap connector?
Claude can query your customer feedback in Unwrap from the chat. Twelve reading tools let it pick a view, discover filters, search themes and segments, draw charts, analyze a sample and consult Unwrap's help docs. Three more create a sample, sort its entries into labeled sub-samples or send feedback to the Unwrap team. The last two tools are not described by any source under their names, so this page leaves them aside.
02Can Claude change my data in Unwrap?
No source describes a tool that edits your customer feedback or your taxonomy. Three tools do produce something, though: a named sample, sub-samples sorted by label, and feedback sent to the Unwrap team about the connector. Unwrap nonetheless states in the same article that all its tools are read-only, and the limits section details that contradiction. A public list stays a floor, since an admin can open more than it shows.
03Does Claude ask before it acts?
By default, yes: Claude asks for confirmation before every action it takes on an account for you. No source describes a confirmation specific to the three tools that create or send something, so the general rule covers them. Unwrap itself recommends leaving reading on permanent allow and writing under approval. On Team and Enterprise, owners can shut writing off for the whole organization, and members cannot lift that from their own accounts.
04Which plans is it available on?
No official source publishes plan availability connector by connector, and no directory sheet displays it. The general rule opens remote connectors to all users on Claude, Cowork, Claude Desktop and mobile. On Team and Enterprise, an Owner or Primary Owner adds the connector to the organization first. For the current state of your own account, check the connector's sheet in the official directory. Connecting then happens from personal settings, Settings > Connectors, as Unwrap explains.
05Does Claude see every view in my organization?
No. It only lists the views your Unwrap permissions give you, and it works with your permissions and nothing more. Unwrap documents one exception: an admin can create shared credentials for the whole organization, and anyone using them then reads every view. If your team picked that option, Claude's access depends on those credentials rather than on your personal account. According to Unwrap, only an admin of the Unwrap organization can create those credentials.
06Why does Unwrap's documentation say read-only?
Unwrap's article states that all its tools are currently read-only, then describes in the same table one tool that creates a named sample, another that creates sub-samples and a third that sends feedback to the Unwrap team. This page follows the tool-by-tool descriptions and places those three on the write side. Nothing suggests they touch your customer feedback itself, and Unwrap advises keeping its write tools under approval in any case.
07Claude or an automation tool for Unwrap?
They do different jobs, so it depends on the use. The connector fits a one-off question asked in a chat: which topics are rising, what customers say about a feature, how it has moved. It launches nothing by itself and watches no theme continuously. An automation tool runs steps without you, starting from an event. The two complement each other more than they compete. With Claude, the question starts with you and the answer lands in the chat.