CommentKeyword

Instagram MCP Server (Comment-to-DM)

An open-source Instagram MCP server that lets Claude, Cursor, Codex and other AI agents set up Instagram comment-to-DM automations ("Comment GUIDE and I'll DM you the link") through Meta's official Instagram API. If you were looking for a ManyChat MCP to build comment automations from Claude, this is the alternative that does it: ManyChat's API can't create keyword or comment triggers, so no ManyChat MCP server can either.

Made by CommentKeyword. MIT-licensed. Registry name com.commentkeyword/mcp.

What it looks like

You:    Edit this into a Reel. CTA: comment GUIDE to get the free checklist.

Claude: …edits the video, adds a "Comment GUIDE 👇" overlay inside the safe zone, writes the caption… Ready to arm the DM: Keyword: GUIDE (whole word) · Posts: your next post or reel DM: "Here's the checklist, {{namefirst}} 👉 https://…" Public reply: "Sent! Check your DMs 📩" Create it? You: Yes Claude: Done. GUIDE is waiting for your next post. Publish the Reel, then comment GUIDE from another account to test it.

The next_post target means you don't need the post's ID: the automation attaches itself to the next post or Reel the account publishes, so you can set it up while the video is still in the edit.

More prompts that work:

• "Anyone who comments PLAN on my next Reel should get https://example.com/plan by DM. Add a public reply."
• "Re-arm GUIDE for my next upload."
• "Which of my keywords are waiting for a post? Pause the ones from last month."
• "Attach my sales agent to the WEBINAR keyword so it answers follow-up questions."

Instagram MCP server vs ManyChat MCP

People often search for a ManyChat MCP when what they want is to tell Claude "set up comment-to-DM for my next Reel". As of October 2026:

ManyChat is still the better tool if you need WhatsApp, Messenger, TikTok or email in one place, or large multi-branch visual flows. Details: ManyChat MCP: what exists and ManyChat alternative for comment-to-DM.

Instagram MCP servers: how this one differs

Meta has no official MCP server for organic Instagram (its official MCP covers ads), so every Instagram MCP server is third-party. Many log in with your password or a session cookie, or scrape public pages. This one only talks to the CommentKeyword API, which uses Meta's official Instagram login and permissions, and it does one job well: comment-to-DM. It doesn't publish posts, read Instagram insights or read your inbox (only the conversations your CommentKeyword AI agents handled); pair it with a scheduler or an analytics connector for those. Comparison: Instagram MCP servers compared.

Connect by URL (no token, no Node.js)

The hosted server at https://commentkeyword.com/connect/mcp has the same tools. Your app signs in with CommentKeyword (OAuth) and you approve what it may do; disconnect it any time in Settings → API & MCP.

• Claude (claude.ai, desktop app, Cowork, mobile): Customize → Connectors → + → add a custom connector, paste the URL, keep the detected options (sign in now, Claude's published identity), then sign in. On Team and Enterprise an owner adds it under Organization settings → Connectors.
• Claude Code: claude mcp add --transport http --scope user commentkeyword https://commentkeyword.com/connect/mcp, then /mcp → commentkeyword → sign in.
• Codex (CLI, IDE extension and app): codex mcp add commentkeyword --url https://commentkeyword.com/connect/mcp opens the sign-in; if it doesn't, run codex mcp login commentkeyword.
• ChatGPT: chatgpt.com/plugins → + → Add custom MCP server, paste the URL, keep OAuth, then sign in (depends on your plan and workspace settings).
• Cursor: { "mcpServers": { "commentkeyword": { "url": "https://commentkeyword.com/connect/mcp" } } } in ~/.cursor/mcp.json, then sign in when Cursor asks.

The rest of this README covers the local server, which uses an API token, and the Claude Code plugin, which also installs the Agent Skill.

Before you start

1. Create a CommentKeyword account at https://commentkeyword.com. API access is included on every plan, including Free.
2. Connect an Instagram professional account (Business or Creator).
3. Create an API token under Settings → API & MCP (https://commentkeyword.com/settings?tab=api). Give it keywords:read, keywords:write and agents:read. Add keywords:delete only if you want the agent to be able to delete automations, and conversations:read if it should read your AI agents' conversations.

Install

Claude Code (skill and MCP server)

/plugin marketplace add commentkeyword/instagram-mcp-server

/plugin install commentkeyword@commentkeyword

Claude Code asks for your API token when the plugin is enabled. It stores the token in your system keychain.

Claude Code (MCP server only)

claude mcp add --transport stdio --env COMMENTKEYWORD_API_TOKEN=YOUR_TOKEN commentkeyword -- npx -y github:commentkeyword/instagram-mcp-server#v0.3.1

Claude Desktop, Cursor, Windsurf and other MCP clients

Add the server to the client's MCP config. For Claude Desktop that file is claude_desktop_config.json (Settings → Developer → Edit Config).

{

"mcpServers": { "commentkeyword": { "command": "npx", "args": ["-y", "github:commentkeyword/instagram-mcp-server#v0.3.1"], "env": { "COMMENTKEYWORD_API_TOKEN": "YOUR_TOKEN" } } } }

Skill only (Claude.ai, Claude Desktop, other agents that read SKILL.md)

With the skills CLI (Claude Code, Cursor, Codex and other agents):

npx skills add commentkeyword/instagram-mcp-server

Or by hand:

1. Download skills/comment-to-dm-cta as a folder and zip it.
2. In Claude, upload the zip under Customize → Skills → Upload a skill.

Without the MCP server, the skill still plans the CTA and the DM. It can also call the REST API directly where the agent has network access (see references/api.md).

Add-ons for video toolkits

n8n

integrations/n8n has an importable n8n workflow that creates the comment-to-DM automation for your next post through the REST API, with Header Auth and no code.

MCP tools

Network and data

• One destination. The MCP server and the skill's curl examples only call the CommentKeyword API, at https://commentkeyword.com/api (or COMMENTKEYWORD_API_URL if you set it). No telemetry, no other hosts.
• What gets sent: your API token (as a Bearer header) and the automation you confirm: keyword, DM and reply text, target posts and options.
• What comes back: your keyword automations and the names of your AI agents.
• Nothing on disk. The server stores nothing locally.
• Privacy policy: https://commentkeyword.com/privacy

Safety

• Every automation messages real people. The skill and the tool descriptions tell the agent to confirm the keyword, the DM and the target with you before creating or activating anything.
• Each token works for one workspace and can be revoked at any time in Settings → API & MCP.
• Use the narrowest permissions you need.

Development

npm test   # end-to-end tests: spawns the server over stdio against a mock API

The MCP server has no dependencies and no build step. It is two readable files in server/: core.mjs holds the tools, the validation and the JSON-RPC handling; `commentkeyw