instagram-mcp

instagram-mcp

An MCP server for the official Instagram Graph API, enabling read, publish, comment, and analytics across Instagram Business/Creator accounts.

Category
Visit Server

README

instagram-mcp

<!-- mcp-name: io.github.adelaidasofia/instagram-mcp -->

An MCP server for the official Instagram Graph API. Read, publish, comment, and pull analytics across one or many Instagram Business/Creator accounts from Claude (or any MCP client) — ToS-safe, no private/reverse-engineered API.

29 tools across accounts, media, publishing, insights, comments, discovery, and (review-gated) direct messages.

npx-style stdio MCP · Python · FastMCP · MIT

Why this one

Most "Instagram automation" tools either (a) wrap the unofficial private API (username + password) — which violates Instagram's Terms and risks a ban — or (b) only post, with no way to read insights, comments, or mentions back. This server is built entirely on the official Graph API, is multi-account from day one, and covers the full read + write surface. The only thing it cannot do without Meta's approval is DMs (see Direct messages).

Requirements

  • An Instagram Professional account (Business or Creator). Personal accounts cannot use the Graph API. Converting is free and reversible (Instagram app → Settings → Account type).
  • The Instagram account connected to a Facebook Page.
  • A long-lived access token with instagram_basic + instagram_content_publish + instagram_manage_comments + instagram_manage_insights, and the account's numeric Instagram Business Account id. See SETUP.md for the exact token + id steps.
  • Python 3.10+.

Install

From PyPI:

pip install adelaidasofia-instagram-mcp

Or from source:

git clone https://github.com/adelaidasofia/instagram-mcp
cd instagram-mcp
python3 -m venv .venv && .venv/bin/pip install -e .

Register it with your MCP client (Claude Desktop / Claude Code) — single-account zero-config path:

{
  "mcpServers": {
    "instagram": {
      "command": "instagram-mcp",
      "env": {
        "INSTAGRAM_MCP_ACCESS_TOKEN": "EAA...your-long-lived-token...",
        "INSTAGRAM_MCP_IG_USER_ID": "17841400000000000"
      }
    }
  }
}

instagram-mcp is the console script the package installs. From a source checkout it lives at .venv/bin/instagram-mcp, or run the module directly with python -m instagram_mcp.server.

Or skip the env vars and call add_account at runtime (token goes to your OS keychain). See Multiple accounts.

Tools

Accounts & healthhealthcheck, list_accounts, add_account, set_default_account, remove_account, account_info

Profile & mediaget_profile, list_media, get_media

Insightsget_account_insights (reach, impressions, profile views, follower count), get_media_insights (per-post reach, saves, shares, interactions), get_audience_insights (follower demographics: age, gender, country, city)

Publishingpublish_image, publish_video, publish_reel, publish_carousel (2–10 images), publish_story, publishing_limit (remaining 24h quota). Video/reel containers process asynchronously; the server polls to completion before publishing.

Commentsget_comments, reply_to_comment, hide_comment, delete_comment

Discoverysearch_hashtag, get_hashtag_media, get_mentions, business_discovery (read any public Professional account by username)

Direct messageslist_conversations, get_messages, send_message (see below)

Publishing takes public https media URLs — Instagram fetches the bytes itself, so the image/video must be reachable on the open web (an S3/Cloudflare/any-CDN URL works).

Multiple accounts

Run one server for all your accounts (yours, a brand's, a client's). Each account authorizes its own token:

add_account(label="brand-a", access_token="EAA...", ig_user_id="178414...", make_default=True)
add_account(label="brand-b", access_token="EAA...", ig_user_id="178414...")
list_media(account="brand-b")
get_account_insights(account="brand-a")

Tokens are stored in the macOS keychain (or a chmod 600 file on other platforms), never in the metadata file and never returned by any tool. Omit account on any tool to use the default.

Direct messages

The DM tools require the instagram_manage_messages permission, which Meta grants only through App Review (typically weeks, and stricter in 2025–2026). Until then, the DM tools fail loud with that instruction rather than silently no-op. Once your app is approved and the token carries the scope, set INSTAGRAM_MCP_DM_ENABLED=1. Note Instagram's 24-hour standard-messaging window applies.

Full App Review walkthrough (screencast shot list, reviewer test-instructions template, privacy-policy + business-verification requirements, post-approval flip, message-tag caveat): docs/APP_REVIEW.md.

Safety

  • Egress is pinned to the Meta host allow-list (graph.facebook.com / graph.instagram.com). A general SSRF guard (RFC1918 / loopback / link-local / CGNAT / cloud-metadata, fail-closed DNS) backs any URL handling.
  • Credentials never leak: every result and error passes a scrubber that strips access tokens (incl. Meta EAA… / IGQV…), bearer headers, app secrets, and API keys before it reaches the model.
  • Observability: every call appends a 4-field JSONL audit line (execution_time_ms, io, token_usage, error_class) under ~/.claude/instagram-mcp/audit.log.jsonl.
  • Input validation runs before every Graph call (ids, caption length, hashtag/username charset, https media URLs).

License

MIT — see LICENSE.

Built by Adelaida Diaz-Roa. Full install or team version at diazroa.com.

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured