Developer Feedback API

The feedback API built for developers

One POST request to collect feedback from your app, server, or AI agent. Query it, search it by meaning, and pipe it anywhere with webhooks. No SDK lock-in, no per-seat pricing.

Get your API key — free Read the API docs

Free tier: 5,000 events/month · Full API access · No credit card

collect-feedback.sh

# Submit feedback from anywhere
curl -X POST https://api.pollenate.dev/collect \
  -H "X-Pollenate-Key: pk_live_..." \
  -d '{
    "inboxKey": "my-app",
    "type": "nps",
    "score": 9,
    "comment": "Love the new dashboard"
  }'

# 201 Created
{ "id": "f3a91c...", "success": true }

Everything is an API

The dashboard is optional. Every capability — collection, querying, search, automations — is exposed through the same API your widgets use.

🔑

Scoped API keys

Mint org-wide or brand-scoped keys with granular scopes — collect, read, write, forms — plus optional expiration dates. Revoke instantly.

📄

REST + OpenAPI 3.1

A clean JSON API described by a published OpenAPI 3.1 spec. Generate typed clients in any language, or explore it with your favorite tooling.

🪝

Webhooks & automations

Fire webhooks on feedback.created and page_response.created with custom headers, body templates, and encrypted secrets interpolation.

🧩

MCP server

Connect Claude, Cursor, or any MCP client to the /mcp endpoint — submit feedback, query stats, and list inboxes without writing HTTP calls.

🔍

Semantic search API

Search feedback by meaning, not just keywords. Vector embeddings let you find "checkout is confusing" when the user wrote "paying is hard".

⚡

Zapier integration

Triggers for new feedback and page responses, plus actions to submit and update feedback — connect 6,000+ apps without code.

🔗

Issue tracker linking

Link feedback to Jira, GitHub, Linear, Asana, or Redmine issues via the API and keep your triage workflow in the tools you already use.

🖥️

Drop-in widget & React

A one-line script tag or a typed React package for in-app collection — both feed the same API and inboxes as your server-side events.

A surface you can learn in an afternoon

Predictable REST endpoints, JSON in and out, documented in the OpenAPI 3.1 spec.

POST/collectSubmit feedback with an API key (widget, server, or agent)

GET/organizations/:orgId/feedbackList + filter feedback by type, score, date, inbox, source

POST/organizations/:orgId/searchSemantic search across all collected feedback

PATCH/organizations/:orgId/feedback/:idUpdate status or link an external issue tracker

GET/v1/organizations/:orgId/feedback-pages/:idRead form structure for programmatic submission

POST/mcpModel Context Protocol endpoint for AI agents

Plus inboxes, brands, feedback pages, notes, automations, and API key management — see the full reference →

AI-native by design

Your agents get their own endpoint

The built-in MCP server lets Claude Desktop, Cursor, and any Model Context Protocol client submit feedback, query stats, and browse inboxes — authenticated with the same scoped API keys.

Explore the AI-native platform →

mcp-config.json

{
  "mcpServers": {
    "pollenate": {
      "url": "https://api.pollenate.dev/mcp",
      "headers": { "X-Pollenate-Key": "pk_live_..." }
    }
  }
}

Frequently Asked Questions

Is there a free tier for the feedback API?

Yes. The Starter plan is free forever and includes full API access, 1 inbox, and 5,000 events per month. No credit card required — create an API key and start posting feedback in minutes.

How does authentication work?

Requests authenticate with an API key sent in the X-Pollenate-Key header. Keys can be scoped to an entire organization or a single brand, restricted by scope (collect, read, write, forms), and given expiration dates.

Is there an OpenAPI specification?

Yes. A complete OpenAPI 3.1 spec is published at pollenate.dev/openapi.yaml covering every public endpoint, auth scheme, and schema — ready for codegen tools and API clients.

Do you support webhooks?

Yes. Automations can deliver webhooks on feedback.created and page_response.created events with custom headers, templated JSON bodies, and encrypted secret interpolation for signing or authorization headers.

Can AI agents submit and query feedback?

Yes. Pollenate ships a Model Context Protocol (MCP) server at /mcp with tools to submit feedback, list inboxes and brands, fetch feedback, and query stats. Claude Desktop, Cursor, and any MCP client connect directly.

What are the rate limits?

Limits are plan-based daily allowances on feedback events and search queries rather than per-second throttles. The API returns clear 429 responses with your limit, plan, and reset time when a cap is reached.

Ship feedback collection today

Start free View pricing