Skip to main content

SentiSift MCP Server

Official Model Context Protocol (MCP) server for SentiSift. Lets MCP-compatible AI hosts (Claude Desktop, Cursor, VS Code, Continue, and others) call the SentiSift comment-moderation and intelligence API as a native tool.

What this unlocks

Once installed, your AI assistant can:

  • Analyze comments directly in conversation: "Run these 50 comments through SentiSift and show me the bots and the sentiment distribution."
  • Check your account state: "What's my SentiSift balance and when does my subscription renew?"
  • Retrieve historical results: "Show me the sentiment history for yesterday's article on example.com/breaking-news."
  • Self-diagnose service status: "Is the SentiSift API ready, or are models still loading?"

No code required. The assistant calls the tools through MCP and reports back.

Tools exposed

Tool Description
analyze_comments Submit a batch of comments plus an article URL, receive bot/spam flags, sentiment labels, commercial-content flags, moderation summary, and (on Pro/Enterprise) Intelligence + interleaved Influence comments.
get_balance Current comment balance, tier, feature flags, active grants, subscription state, Influence stats.
get_health Whether the SentiSift API is ready or still loading models.
get_article_results Historical scored comments for an article URL (Pro/Enterprise retain full history; Free/Starter return buffer state only).

Install

You need Python 3.10+ and a SentiSift API key. Free keys (1,000 comments, no credit card) at sentisift.com/pricing.

uv runs published Python apps without persistent install. This is the install-free path.

# One-time: install uv (mac/linux)
curl -LsSf https://astral.sh/uv/install.sh | sh

# Verify it works (Ctrl-C after a few seconds; the server is meant to be
# launched by your AI host, not run in a terminal).
SENTISIFT_API_KEY=sk_sentisift_... uvx sentisift-mcp

Alternative: pip install

pip install sentisift-mcp
sentisift-mcp   # run manually to verify

Host configuration

Each AI host has its own config file. Pick yours:

Claude Desktop

Edit the config file:

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json

Add the SentiSift entry to mcpServers:

{
  "mcpServers": {
    "sentisift": {
      "command": "uvx",
      "args": ["sentisift-mcp"],
      "env": {
        "SENTISIFT_API_KEY": "sk_sentisift_your_key_here"
      }
    }
  }
}

Restart Claude Desktop. You should see a new tool indicator in the bottom-left corner of a new chat. Try: "Use SentiSift to check my balance."

Cursor

Cursor supports MCP via the Settings → MCP Servers panel. Add:

  • Name: sentisift
  • Command: uvx
  • Arguments: sentisift-mcp
  • Environment variables: SENTISIFT_API_KEY=sk_sentisift_your_key_here

Or edit ~/.cursor/mcp.json directly with the same structure as the Claude Desktop example above.

VS Code (with the MCP extension)

Edit your user or workspace settings.json:

{
  "mcp.servers": {
    "sentisift": {
      "command": "uvx",
      "args": ["sentisift-mcp"],
      "env": {
        "SENTISIFT_API_KEY": "sk_sentisift_your_key_here"
      }
    }
  }
}

Continue.dev

In ~/.continue/config.json:

{
  "mcpServers": [
    {
      "name": "sentisift",
      "command": "uvx",
      "args": ["sentisift-mcp"],
      "env": {
        "SENTISIFT_API_KEY": "sk_sentisift_your_key_here"
      }
    }
  ]
}

Any other MCP-compatible host

The server speaks MCP over stdio. Point your host at uvx sentisift-mcp (or sentisift-mcp if installed via pip) and set SENTISIFT_API_KEY in its environment.

How to talk to the assistant

Once configured, natural-language prompts work. Some examples:

"Check my SentiSift balance."

"Run these 20 comments through SentiSift for the article at https://example.com/breaking-story. Here are the comments:" (paste them)

"Is the SentiSift API ready right now, or is it still loading?"

"Fetch the processed comments for https://example.com/article-42 from SentiSift."

The assistant chooses the right tool, passes the right arguments, and summarizes the results. If a call fails, the error response includes a deep link to the relevant docs section so the assistant can self-correct.

Privacy

  • The server sends a User-Agent header identifying it as sentisift-mcp/<version> so we can distinguish MCP traffic from direct SDK traffic in our usage logs. No other telemetry.
  • The server runs locally on your machine. It only calls out to api.sentisift.com.
  • Your API key stays on your machine (in the host's env config). It is never logged to disk by this server.

Source

Source and issue tracker: gitlab.com/pickel-fintech/sentisift-sdks.

License

MIT — see LICENSE. The SentiSift API service that this client calls is governed separately by sentisift.com/terms.html; use of the API requires an API key from sentisift.com/pricing.

Metadata

Release files for sentisift-mcp 0.1.4

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for sentisift-mcp 0.1.4
File Size Uploaded Provenance
sentisift_mcp-0.1.4.tar.gz 9.8 kB Without provenance Details

Built distribution (wheel)

Table of built distributions (wheels) for sentisift-mcp 0.1.4
File Interpreter ABI Platform
sentisift_mcp-0.1.4-py3-none-any.whl Python 3 none any Details

Total release size: 19.7 kB

Release files / sentisift_mcp-0.1.4.tar.gz

Download URL sentisift_mcp-0.1.4.tar.gz
Size 9.8 kB
Tags Source
SHA-256 checksum
How to use checksums
5c7fbcc994a4d8691daa927102065a214a00ac2c857728a6f5a21f7c28a50830
BLAKE2b-256 checksum
How to use checksums
6f1814d32e1fe9323f3d312a5f49184bf91f5d36675cd34c12e1ace380577f11
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.11.17

Release files / sentisift_mcp-0.1.4-py3-none-any.whl

Download URL sentisift_mcp-0.1.4-py3-none-any.whl
Size 9.9 kB
Tags Python 3
SHA-256 checksum
How to use checksums
4024bbebd75272016f3b0ba7005a3708c982dbbc9404ad70bc739e9d2000f69c
BLAKE2b-256 checksum
How to use checksums
618b8ccaeb62fe60ee0c072793f43ca71973e6d323e6bdeea5887fb2408e2f57
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.11.17

Security for sentisift-mcp 0.1.4

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.1.4 This release

2 release files No provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page