Sprout Social MCP Server
This server acts as a Model Context Protocol (MCP) interface, enabling AI assistants to interact with Sprout Social data for analytics, publishing, messaging, social listening, media management, and customer cases.
Customer Metadata: Retrieve customer IDs, social profiles, groups, tags, users, teams, listening topics, and case queues.
Analytics: Get profile-level metrics (impressions, engagements) and post-level metrics (impressions, engagements, reactions, video views) for specific date ranges, with pagination support.
Messages: Fetch inbox messages with filtering by profile, date range, and cursor-based pagination.
Social Listening: Access metrics (volume, sentiment, engagement) and messages for specific listening topics.
Publishing: Create and schedule posts to one or more social profiles, with support for drafts, tags, and media attachments. Retrieve details of existing posts by ID.
Media Management: Upload images or videos via a public URL for use in publishing posts.
Case Management: Retrieve customer cases/inquiries with filters for priority, time range, and pagination.
Prerequisites: Requires a Sprout Social API Key and Customer ID. Compatible with AI assistants such as Claude Desktop, Cursor, Devin, and VS Code (GitHub Copilot).
Allows managing Facebook profiles, posts, and analytics (impressions, engagements, reactions, etc.) through the Sprout Social API.
Allows managing Instagram profiles, posts, and analytics (impressions, engagements, reactions, video views, etc.) through the Sprout Social API.
Allows managing TikTok profiles, posts, and analytics (impressions, engagements, reactions, video views, etc.) through the Sprout Social API.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Sprout Social MCP Server get my Instagram analytics for last week"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Sprout Social MCP Server
Note: This is an unofficial, community-built MCP server to use while Sprout Social works on releasing their official one.
A Model Context Protocol (MCP) server for the Sprout Social API. It lets AI assistants (Claude, Cursor, Devin, etc.) access your Sprout Social data — analytics, publishing, messages, listening, and more — through a standardized interface.
Quick Start
Prerequisites
Node.js 18+
A Sprout Social API token (how to create one)
Your Sprout Social Customer ID (how to find it)
Running via npx
No installation required:
SPROUT_SOCIAL_API_KEY=your-token \
SPROUT_SOCIAL_CUSTOMER_ID=your-customer-id \
npx sprout-social-mcpConfiguration with Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"sprout-social": {
"command": "npx",
"args": ["-y", "sprout-social-mcp"],
"env": {
"SPROUT_SOCIAL_API_KEY": "your-api-token",
"SPROUT_SOCIAL_CUSTOMER_ID": "your-customer-id"
}
}
}
}Configuration with Cursor
Add to your .cursor/mcp.json:
{
"mcpServers": {
"sprout-social": {
"command": "npx",
"args": ["-y", "sprout-social-mcp"],
"env": {
"SPROUT_SOCIAL_API_KEY": "your-api-token",
"SPROUT_SOCIAL_CUSTOMER_ID": "your-customer-id"
}
}
}
}Configuration with VS Code (GitHub Copilot)
Add to your .vscode/mcp.json:
{
"servers": {
"sprout-social": {
"command": "npx",
"args": ["-y", "sprout-social-mcp"],
"env": {
"SPROUT_SOCIAL_API_KEY": "your-api-token",
"SPROUT_SOCIAL_CUSTOMER_ID": "your-customer-id"
}
}
}
}Configuration with Devin
In Devin's MCP settings, add a new server:
Name:
sprout-socialCommand:
npx -y sprout-social-mcpEnvironment Variables:
SPROUT_SOCIAL_API_KEY→ your API tokenSPROUT_SOCIAL_CUSTOMER_ID→ your customer ID
Related MCP server: sprout-mcp
Environment Variables
Variable | Required | Description |
| Yes | Your Sprout Social API token |
| If the token sees more than one customer | Your Sprout Social customer ID. Optional when the token has access to exactly one customer — the server reads it from |
Available Tools
Customer Metadata
Tool | Description |
| Get your Sprout Social customer IDs and names |
| List all connected social profiles |
| List all groups |
| List tags, with optional filters for active status, group, type, or name |
| List all users |
| List all listening topics |
| List all teams |
| List all case queues |
Analytics
Tool | Description |
| Profile-level analytics (impressions, engagements, etc.) for a date range |
| Post-level analytics with pagination. Supports impressions, engagements, reactions, video views, and tag filters |
| Tag Performance Report-style rollup: automatically pull tagged posts and aggregate lifetime metrics by tag |
Messages
Tool | Description |
| Retrieve inbox messages with filtering and cursor-based pagination |
Listening
Tool | Description |
| Get metrics for a listening topic |
| Get messages from a listening topic |
Publishing
Tool | Description |
| Create a new post to be published at a scheduled time |
| Retrieve details of a specific publishing post |
Media
Tool | Description |
| Upload media via URL for use in publishing posts |
Cases
Tool | Description |
| Retrieve customer cases/inquiries with filters for priority, time range, etc. |
Usage Tips
Post Analytics Pagination
The Sprout Social API paginates post analytics (50 posts per page). get_post_analytics can follow that for you:
all_pages: true— fetch pages automatically (default capmax_pages: 40) and return one combineddataarray.paging.truncatedis true if more pages remain.page— fetch a single page yourself.guid_cursor— walk past the ~10k page-number cap. Pass the lastguid(orpaging.next_guid_cursor) and keep going until a page is empty. Combine withall_pagesto follow the cursor automatically.
sort (e.g. ['lifetime.impressions:desc']) and timezone (ICANN name for the date filter; response times stay UTC) are also supported. Cursor mode always sorts by guid:asc. Requests retry on 429/5xx and pause briefly between pages.
Ask: "Get all Instagram post analytics for last week"
→ get_post_analytics with all_pages=true
Ask: "Top posts by impressions last month"
→ get_post_analytics with sort=['lifetime.impressions:desc']Valid Post-Level Metrics
All platforms (Instagram, Facebook, LinkedIn, TikTok, etc.):
lifetime.impressions— total viewslifetime.engagements— total engagement (likes, comments, shares, saves)lifetime.reactions— reactions onlylifetime.video_views— video view countlifetime.saves— saves/bookmarkslifetime.comments_count— comment countlifetime.post_shares_count— share count
Facebook only:
lifetime.post_link_clicks— clicks on links in the postlifetime.post_content_clicks— total clicks on post contentlifetime.post_content_clicks_other— other content clicks
Platform limitations:
Reach (
lifetime.reach) is NOT available at the post level — only at the profile level viaget_profile_analytics. This is a Sprout Social API limitation.Click metrics are only available for Facebook posts. For Instagram, the API silently ignores them (no error, but no data returned).
Invalid metrics (will cause errors): lifetime.comments, lifetime.shares, lifetime.reach
Tag Performance (replaces exporting a tag report)
Sprout has no dedicated tag-analytics endpoint. Tag data lives on posts (internal.tags.id) and can be combined with post metrics to reproduce the Tag Performance Report.
get_tag_performance does that automatically:
Resolves tag names/IDs via
get_tagsmetadataPages through tagged posts in the date range
Aggregates lifetime metrics per tag (impressions, engagements, engagement rate, reactions, comments, shares, saves, video views, Facebook link clicks)
Returns top posts for each tag, and optionally a per-network breakdown
Ask: "How did the campaign and launch tags perform on Instagram last month?"
→ get_profiles (for Instagram customer_profile_id)
→ get_tag_performance with tag_names ["campaign", "launch"]Behavior notes:
A post with multiple tags contributes its full lifetime metrics to each tag — the same method Sprout uses in the Tag Performance Report.
If
tag_ids/tag_namesare omitted, every tagged post in the window is rolled up.Ambiguous tag names return candidate IDs instead of guessing. Use
get_tags(search,active_only,group_id) to pick the right ID.Stories often inflate post counts and dilute engagement rate. Pass
exclude_post_types: ["INSTAGRAM_STORY"]when you want feed/reel performance only.Responses set
truncated: trueifmax_pagesstopped before all post pages were fetched (50 posts per page).
You can also filter raw posts without aggregating:
get_post_analytics(..., tag_ids: ["12345"], tagged_only: true)The filter uses internal.tags.id.eq(...) / internal.tags.id.exists(true) on the Posts Analytics API. tag_id.eq(...) is not valid on that endpoint.
Finding Profile IDs
Use get_profiles first to discover your customer_profile_id values, then pass them to analytics or publishing tools.
Development
git clone https://github.com/jginorio/sprout-social-mcp.git
cd sprout-social-mcp
npm install
npm run build
npm testTo test locally:
SPROUT_SOCIAL_API_KEY=your-token \
SPROUT_SOCIAL_CUSTOMER_ID=your-customer-id \
node dist/index.jsLicense
MIT
Available Tools
18 toolscreate_publishing_postA
Create a new publishing post in Sprout Social to be published at a future time. The post will appear in Sprout's publishing calendar.
Inputs: text (required), profile_ids (required), scheduled_time (required), tags, +2 more.
get_case_queuesA
List all case queues in your Sprout Social account.
Inputs: no named parameters.
get_casesB
Retrieve cases (customer inquiries/issues) from Sprout Social. Cases represent customer interactions that may require action by a social care agent.
Inputs: sort, limit, priority, timezone, +3 more.
get_clientB
Get your Sprout Social customer IDs and names.
Inputs: no named parameters.
get_groupsA
List all groups in your Sprout Social account.
Inputs: no named parameters.
get_listening_topic_messagesA
Get messages found within a specific listening topic. Use get_topics first to discover available topic IDs.
Inputs: topic_id (required), page, sort, limit, +2 more.
get_listening_topic_metricsA
Get metrics for a specific listening topic (e.g. volume, sentiment, engagement). Use get_topics first to discover available topic IDs.
Inputs: metrics (required), topic_id (required), page, filters.
get_messagesA
Retrieve messages from your Sprout Social inbox. Supports cursor-based pagination. Messages include those received by and sent from your profiles.
Inputs: profile_ids (required), sort, limit, fields, +2 more.
get_post_analyticsA
Get post-level analytics (impressions, engagements, etc.) for posts within a date range. Supports filtering by Sprout tags via tag_ids or tagged_only. Responses include internal.tags.id by default so posts can be grouped by tag. Supports sort (e.g. lifetime.impressions:desc), timezone, page, all_pages auto-pagination, and guid_cursor for walking past the ~10k page cap. For a Tag Performance Report-style rollup, prefer get_tag_performance. IMPORTANT: The page parameter must be in the request body, not as a URL query parameter.
Inputs: metrics (required), profile_ids (required), created_time_end (required), created_time_start (required), +10 more.
get_profile_analyticsA
Get owned profile-level analytics (impressions, engagements, etc.) for one or more profiles over a reporting period. Requires a reporting_period filter in the format 'YYYY-MM-DD...YYYY-MM-DD' (max 1 year span).
Inputs: metrics (required), profile_ids (required), reporting_period_end (required), reporting_period_start (required), +1 more.
get_profilesA
List all social profiles connected to your Sprout Social account.
Inputs: no named parameters.
get_publishing_postB
Retrieve details of a specific publishing post by its ID.
Inputs: publishing_post_id (required).
get_tag_performanceA
Automatically pull Tag Performance Report-style analytics: fetch tagged posts in a date range, resolve tag names, and aggregate lifetime metrics by tag (and optionally by network). Replaces exporting a tag report from Sprout for post insights. Use get_tags first if you need to discover tag names/IDs. If tag_ids and tag_names are omitted, every tagged post in the window is rolled up. A post with multiple tags contributes its full metrics to each tag. Pages through the Posts Analytics API automatically.
Inputs: profile_ids (required), created_time_end (required), created_time_start (required), metrics, +7 more.
get_tagsA
List message/post tags in your Sprout Social account. Optionally filter by active status, group, type (LABEL or CAMPAIGN), or name search. Use these tag_id values with get_tag_performance and get_post_analytics.
Inputs: type, search, group_id, active_only.
get_teamsA
List all teams in your Sprout Social account.
Inputs: no named parameters.
get_topicsA
List all listening topics in your Sprout Social account.
Inputs: no named parameters.
get_usersA
List all users in your Sprout Social account.
Inputs: no named parameters.
upload_mediaB
Upload media (image/video) to Sprout Social for use in publishing posts. Provide either a public URL to the media file. Returns a media ID to use with create_publishing_post.
Inputs: media_url (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 18 tools
Most tools are clearly distinct by resource (clients, profiles, groups, users, topics, teams, cases, tags, messages, posts, media). However, get_post_analytics and get_tag_performance both deal with post analytics and could be confused, though descriptions clarify their different aggregation purposes.
The vast majority follow a consistent get_<resource> or create_/upload_ verb_noun pattern. Minor deviations like get_listening_topic_messages vs get_messages and get_profile_analytics vs get_post_analytics are still readable and consistent in style.
18 tools is on the higher end but appropriate for a broad social media management platform covering analytics, publishing, listening, and care. Each tool maps to a distinct functional area, though a few could be consolidated.
The surface covers core read operations across major Sprout Social domains plus publishing creation and media upload. Gaps include no update/delete for publishing posts, no case management actions, and no profile management, but the main analytics and listening workflows are well covered.
Maintenance
Related MCP Connectors
MCP server that lets AI assistants use all OneSchema features exposed via the public API.
MCP server connecting AI agents to 100+ apps (Gmail, Slack, Notion, GitHub) via one-click OAuth.
MCP server for building and testing AI agents with multi-model experimentation and insights.
Connect any AI agent to 11+ social platforms: schedule, publish & track posts via hosted MCP.
Related MCP Servers
- AlicenseDqualityDmaintenanceAn MCP server that allows AI assistants to interact with the ServiceTitan API, requiring client credentials for authentication.4ISC
- FlicenseAqualityDmaintenanceAn MCP server that integrates the Sprout Social Public API with Claude and other MCP clients. It allows users to manage social profiles, track analytics for posts and profiles, and schedule social media content.11-
- AlicenseAqualityCmaintenanceMCP server that enables AI assistants to query Parse.ly analytics for top posts, authors, tags, referrers, content search, and social shares.619 npm1MIT
- AlicenseNot gradedqualityDmaintenanceA unified MCP server that gives AI assistants direct access to Google Analytics 4, Google Search Console, Google Ads and Google Merchant Center data using OAuth2 user credentials.MIT