Let your AI app manage your testimonials
Add the KikoPraise connector to Claude, Cursor, Gemini or Windsurf once, then ask in plain English: “add a 5-star testimonial from Grace Hopper, and approve it.”
create_testimonialset_testimonial_statusAPPROVEDDone. I added Grace Hopper’s 5-star testimonial and approved it, so it will show on any widget that includes it.
KikoPraise halved our review collection time.
Connected in a minute
- 1
Create a key
In the dashboard, open MCP connector and create a connection. Choose read only or read and write. The key is shown once.
- 2
Paste it into your AI app
The page gives you the setup for your app. The connector page shows Connected as soon as it sees the first request.
- 3
Just ask
“How many testimonials are pending?” “Add one from Grace Hopper.” The app calls KikoPraise for you.
Nine tools, read or write
Read-only keys can't see or call the write tools.
get_workspace_summaryReadCounts by status, average approved rating, top tagslist_testimonialsReadFilter by status, type, tag, rating or searchget_testimonialReadOne testimonial, with its transcriptlist_tagsReadTags in use, with countslist_formsReadCollection forms and their share linkslist_widgetsReadWidgets and public page linkscreate_testimonialWriteAdds a text testimonial. Pending by defaultupdate_testimonialWriteEdit author, company, content, rating or tagsset_testimonial_statusWritePending, Approved or RejectedKept deliberately small
Scoped to one workspace
Every call is limited to the workspace the key belongs to. Only a hash of the key is stored, and revoking it takes effect on the next request.
New praise is Pending
A testimonial an AI adds is Pending unless you ask it to approve. It is idempotent on author and text, so a retry doesn't duplicate it.
No deleting
There is no tool to delete testimonials, edit forms or widgets, or upload files. An AI app can add and approve praise; it can't remove it.
Standard MCP
Streamable HTTP with JSON responses, stateless. Any client that speaks MCP can use it, not only the ones named here.
Questions about this
What is MCP?
Model Context Protocol, a standard plug that lets AI apps use other software. You install KikoPraise's connector once in an AI app, then talk to that app normally and it calls KikoPraise for you.
Do I need to write code?
No. You paste a key and a URL into your AI app. The REST API and webhooks are there if you do want to write code.
Can the AI publish praise on its own?
Only if you tell it to approve, and only with a read and write key. Otherwise what it adds waits as Pending for you.
Which AI apps can I use it with?
Any app that speaks MCP, such as Claude, Cursor, Gemini or Windsurf. The dashboard page gives you the setup for the app you pick.
How is my workspace protected?
Every call is scoped to the workspace the key belongs to. Only a hash of the key is stored, it is shown once, and revoking it takes effect on the next request.
Ready to collect some love?
Ten minutes to your first story live on your own site.
Nothing publishes until you approve it