EchoThread

What the EchoThread MCP server does

The EchoThread MCP server exposes comment-management functions for sites using EchoThread, a comment system for blogs and documentation. It is intended for use from Claude, ChatGPT, or another MCP client that can connect to the published package.

The available workflow centers on reviewing and moderating comments. An agent can read the pending queue, approve a comment, reject it, mark it as spam, or send a reply. These operations cover the comment actions identified in the project description; the supplied material does not document additional search, analytics, or site-administration functions.

The repository also contains a separate n8n connector with a new-comment trigger and moderation actions. That package is distinct from the MCP server and should not be treated as part of the MCP tool set.

How it works

The EchoThread MCP server communicates with the EchoThread public API. A user creates an API token at EchoThread’s API-token page and provides that token to the connector according to the package’s configuration instructions. The token determines which operations are permitted: an action is available only when its required scope is present.

This scope-based model matters when configuring an agent with moderation access. A token can be restricted so that the client receives only the permissions appropriate for its job. The README does not list the individual scope names or describe how the MCP transport is configured, so those details should be verified in the package documentation before deployment.

Setup and configuration

Create an API token through echothread.io/api-tokens before connecting the server. The token is used to authenticate requests to the EchoThread public API, and its scopes control what the connector can do.

The provided material identifies the npm package as @echothread/mcp and states that releases are published to npm with provenance. It does not provide an install command, a required environment-variable name, a configuration file format, or a client-specific setup example. Use the package documentation or source repository for those implementation details.

Tools and capabilities

The documented capabilities are:

• Read the pending comment queue.
• Approve comments.
• Reject comments.
• Mark comments as spam.
• Reply to comments.

These actions are designed for interactive or agent-assisted comment moderation. The project description does not specify tool names, argument schemas, response formats, pagination behavior, or whether actions can be combined into one operation.

Limitations and notes

The EchoThread MCP server depends on access to the EchoThread public API and a valid token with suitable scopes. A token without the required permission cannot perform the corresponding moderation action.

The supplied README does not document rate limits, retry behavior, error responses, comment-site discovery, or transport support. It also names Claude and ChatGPT as usage contexts but does not provide client-specific configuration steps. Confirm those details in the package documentation before relying on the connector for unattended moderation.

The repository is MIT licensed and publishes the MCP package under the registry identifier io.echothread/mcp.