built by solve-studio

Privacy·Terms

LOOP
DocsSign in
API-first feedback platform

Feedback infrastructure
your whole stack can talk to

A REST API, MCP server, and embeddable widget. Loop lets your team, your clients, and your AI agents create, triage, and resolve feedback programmatically.

Read the docsStart your free trial
01

Install the widget

One script tag. Loads async, zero performance impact. Your users annotate the live page — screenshots, screen recordings, and pins included.

$<script src="loop.js" />
02

Connect the API

Full CRUD over projects, feedback, comments, boards, and clients. Authenticate with an org-scoped API key and plug into any stack.

$GET /api/v1/feedback
03

Let agents take over

AI agents read feedback, create tickets, and triage issues through the MCP server — natively inside their workflow.

$npx @loop-feedback/mcp
> primitives

Built API-first.
Designed for automation.

Every entity in Loop is accessible through the REST API. The MCP server gives AI agents native read/write access. Pipe feedback into your workflow — or let your agents close the loop for you.

API

Public REST API

Full CRUD over projects, feedback items, comments, boards, and clients. Authenticate with an org-scoped API key. JSON everywhere, predictable pagination, standard HTTP status codes.

  • Projects, boards, clients
  • Feedback items & comments
  • Labels, statuses, assignments
  • Org-scoped API key auth
# List feedback for a project
curl -H "Authorization: Bearer $API_KEY" \
https://loop.solve-studio.co/api/v1/feedback
# Response
{
"data": [{
"id": "fb_9x2k...",
"title": "Hero image overlaps nav",
"type": "bug",
"status": "open"
}]
}
# Add to your MCP config
{
"mcpServers": {
"loop-feedback": {
"command": "npx",
"args": ["@loop-feedback/mcp"]
}
}
}
MCP

MCP Server for AI agents

A Model Context Protocol server that lets coding agents read feedback, create tickets, manage boards, and triage issues — natively inside Cursor, Windsurf, or any MCP-compatible tool.

  • One-line install with npx
  • Read & write feedback items
  • Manage boards and columns
  • Auto-triage into bug / feature / task
AI

Agentic ticket creation

AI agents can programmatically create, categorise, and assign feedback items. Close the loop between your codebase and your users — no human sorting required.

Combine with Gmail and Slack integrations to funnel every input channel into a single board. AI triages each item as a feature, bug, or task automatically.

# Agent creates a typed ticket
curl -X POST \
-H "Authorization: Bearer $API_KEY" \
-d '{
"title": "Add dark mode toggle",
"type": "feature",
"description": "Users requesting..."
}' \
https://loop.solve-studio.co/api/v1/feedback
Integrations

Every channel feeds
the same board.

Widget annotations, Gmail threads, and Slack conversations all become typed cards on your kanban board. AI triages everything — you just ship.

widget

Embeddable Widget

One script tag. Clients annotate the live site, capture screenshots, and record their screen — context included.

gmail

Gmail Sync

Label-based mapping turns email threads into board cards. AI reads each thread and classifies it as feature, bug, or task.

slack

Slack Commands

/loop-feedback opens a form. /card on any thread creates a board card. Screenshots, videos, and voice notes supported.

ai

AI Triage

Automatic classification into feature, bug, or task. Extracts multiple cards from a single thread. No manual sorting.

Feedback infrastructure your stack can talk to. Ship the widget, connect the API, let agents do the rest.

Read the docsStart your free trial