Project Echo

Project Echo MCP server

Your customer feedback, in every AI tool you use

Connect Project Echo to Claude, Cursor, and other MCP clients. Ask questions about requests, turn them into implementation prompts for AI, take action from chat, and ground agent workflows in real customer signal — included on the free plan.

Sign up for free

No credit card required

Ask about your feedback

List requests by board or status, inspect comments and votes, and pull the context your agent needs before it drafts a PRD or summary.

Turn requests into implementation prompts for AI

Generate Anthropic Claude Sonnet briefs from a request’s title, description, comments, and internal notes — then fetch or regenerate them via MCP for Cursor and other coding agents.

Take action from your LLM

Create requests, add comments, cast votes, and write internal notes without leaving Claude, Cursor, or another MCP-compatible client.

Ground AI workflows in real data

Use Project Echo as the customer-signal layer for agent workflows — triage feedback, prioritize next work, and keep product context in sync.

Permissioned by design

API keys are scoped with granular permissions and presets like Read-only, Contributor, or Full agent — so agents only get the access you intend.

What agents can do today

The MCP server exposes focused tools over your workspace — the same permissioned surface as the REST API.

Requests & boards

  • List and create requests
  • Filter by board or status
  • List boards and statuses
  • Generate and get implementation prompts

Conversation

  • Create public comments
  • Create votes
  • Read and write internal notes

People

  • List portal members
  • List workspace teammates
  • Update teammate profiles

Up and running in minutes

Create a key, connect your client, and start asking about feedback.

  1. 1

    Create an API key

    In your workspace, open Settings → Connections → API & MCP. Create a key with the permissions your agent should have.

  2. 2

    Add the MCP server

    Point your MCP client at @project-echo/mcp with your API base URL, key ID, and token. Most clients are ready in a couple of minutes.

  3. 3

    Ask and act

    Start with “What are the top open requests on our feedback board?” then generate an implementation prompt for a request and hand it to your coding agent.

Example MCP client config
{
  "mcpServers": {
    "project-echo": {
      "command": "npx",
      "args": ["-y", "@project-echo/mcp"],
      "env": {
        "PE_API_BASE_URL": "https://api.projectecho.io",
        "PE_API_KEY_ID": "<key-uuid>",
        "PE_API_TOKEN": "pe_api_..."
      }
    }
  }
}

Put your agents to work on real customer feedback

Sign up free, create an API key, and connect MCP in a few minutes. Pair it with your feedback portal, roadmap, and changelog for the full loop.

Sign up for free

Frequently asked questions