Pricing
Summarize pricing-page feedback from the last 30 days
Ask an agent to find recent pricing concerns, group repeated objections, and suggest the next pricing-page experiment.
Ask ChatGPT and Claude what your customers are saying
Bring feedback collected in Signaldeck into your AI conversations through MCP. Explore pricing objections, onboarding problems, and feature requests without copying responses between tools.
Read-only access to your Signaldeck responses and available summaries.
In practice
Ask AI tools what users said without copy-pasting feedback
A customer feedback MCP server is a Model Context Protocol endpoint that lets an AI tool retrieve feedback context from a product system. Signaldeck's server exposes supported, account-scoped retrieval tools for feedback already collected in Signaldeck.
Ask ChatGPT, Claude, or another MCP-compatible tool what users said about pricing, onboarding, reported issues, churn reasons, feature requests, or the latest submissions.
Collect customer feedback with the widget to give your AI tools a source of customer responses.
- Pricing objections
- Onboarding friction
- Bug reports
- Churn reasons
- Feature requests
- Recent submissions
Example workflows
Use recent feedback inside the product work you already do
Give agents customer evidence before they write, plan, or route work. Build an AI-native improvement loop from feedback to issue draft, roadmap note, weekly readout, or product decision.
Retention
Find cancellation reasons before changing the flow
Pull in churn and cancellation feedback before you simplify the flow, test retention prompts, or change the questions in the cancellation experience.
Issue tracker
Draft a product issue from repeated onboarding friction
Turn repeated feedback into a well-sourced issue draft. Signaldeck supplies the customer context; your agent can create the issue.
Claude Code
Ask Claude Code what feedback says before changing a feature
Bring customer context into the coding session so implementation choices reflect what users actually struggled with.
Weekly readout
Prepare a weekly feedback readout from Signaldeck summaries
Use summaries and response search to brief the team on emerging themes, risks, positive signals, and decisions that need an owner.
Give agents a direct path to the feedback they need
Signaldeck's MCP tools let agents find the right project, inspect forms and submissions, search response text, and read current summaries through one consistent feedback source.
projects.list
Find available feedback projects
List the Signaldeck projects the connected account can read.
forms.list
See feedback forms in a project
Find the forms tied to a product area, page, workflow, or research question.
submissions.list
Review recent submissions
Pull recent feedback for a project or form, with date filters when the agent needs a specific window.
submissions.get
Open one submission with its answers
Inspect the source response behind a theme, issue, or summary before acting on it.
responses.search
Search feedback text
Find responses that match a phrase, product area, objection, bug report, or customer question.
insights.summary
Read the current feedback summary when available
Start from an available Signaldeck readout before drafting a roadmap note, experiment, or team update.
Bring customer evidence into AI workflows with focused, read-only access
Add Signaldeck’s remote server to an MCP-compatible client, sign in and authorize the right account, then ask the client to retrieve relevant feedback. Access is read-only and account-scoped, with access tokens that expire after 90 days.
Connect the Signaldeck MCP server using the server URL, OAuth instructions, and tool reference in the setup guide.
Use this server URL in your MCP-compatible client.
- OAuth authorization
- Account-scoped feedback
- Supported tools only
- 90-day access-token expiry
- No manual copy-paste
Give agents customer evidence, not loose notes
Signaldeck gives agents a consistent source for customer evidence through projects, forms, submissions, response search, and available summaries, all connected to feedback collected inside Signaldeck.
Questions about the Signaldeck MCP server
What is a customer feedback MCP server?
A customer feedback MCP server is a Model Context Protocol endpoint that lets AI tools retrieve feedback context from a product system. Signaldeck's MCP server lets agents access feedback collected on your own site through the Signaldeck feedback widget, including submissions, response search results, and summaries.
Does it work with ChatGPT?
Yes. Signaldeck can be set up as a ChatGPT app by creating a connector for the Signaldeck MCP server, then choosing the app from the tools menu in a ChatGPT conversation.
Does it work with Claude?
Yes. In Claude, add Signaldeck as a custom connector for the remote MCP server. In Claude Code, add the Signaldeck MCP server as a remote HTTP server and complete the sign-in flow when Claude asks for it.
What Signaldeck data can agents access?
Connected tools can use supported read tools for projects, forms, submissions, response search, and Signaldeck summaries that the connected account can access.
Can connected AI tools change my Signaldeck data?
No. The MCP tools retrieve feedback context for analysis. They do not modify your projects, feedback widget, collected feedback, or summaries.
How is OAuth handled?
After an MCP client starts the connection, Signaldeck asks the user to sign in and authorize MCP access. The client receives a time-limited OAuth token for the read scope and can call the supported MCP tools until the token expires.
How is this different from copying feedback into ChatGPT?
Copying feedback into a chat gives the model a static bundle of raw customer text. MCP lets a connected tool retrieve scoped Signaldeck context when it needs it, including summaries and source submissions, without manually pasting feedback into every conversation.
Go deeper on the MCP server
Educational workflow
Use current feedback inside an agent-assisted decision loop
See how a team can move from a focused question to fresh feedback context, a small action, and a follow-up question.
Create a feedback loop with MCPRelated product pages
Keep following the feedback loop
- See how AI feedback analysis works
AI feedback analysis
AI feedback analysis for themes, risks, and next steps
Use AI to surface patterns and suggested actions while keeping responses visible.
- See product feedback workflows
In-app feedback for SaaS
Collect product feedback while the experience is fresh
Ask users about the product moment they just experienced, before context fades.
- See website feedback use cases
On-page surveys for website feedback
Find out why website visitors hesitate
Ask visitors why they hesitated, what they expected, and what would reduce risk.
Bring current customer evidence into your next AI-assisted workflow
Connect the MCP server to explore the Signaldeck responses you already have, or start collecting focused feedback your AI tools can use next.