Skip to main content
Octolens MCP lets AI assistants like Claude, ChatGPT, and Cursor access your Octolens workspace directly. Instead of switching between tabs and dashboards, just ask questions in natural language — your AI assistant browses mentions, filters by source or sentiment, runs analytics, and manages your monitoring setup on your behalf. MCP (Model Context Protocol) is an open standard that connects AI tools to third-party services in a secure, consistent way.
For shell scripts, CI pipelines, and bulk export, the Octolens CLI is often the better surface — same data, stable exit codes, and a strict --json contract built for automation.
  • “Generate a share of voice graph using my BRAND, COMPETITOR 1, and COMPETITOR 2, with percentage share, mention volume, and sentiment breakdown.”
  • “Show me all Reddit mentions from the last week where sentiment is negative.”
  • “Add ‘next.js deployment’ as a new keyword and route buy-intent mentions to my #leads Slack channel.”
  • “Summarize the feature requests from the last 7 days.”

​
What’s new in v2

v2 exposes the full Octolens surface — every action available in the app is available to your AI assistant. Reads: mentions, keywords, feeds, tags, keyword suggestions, workspace and company profile, usage, and global filters, plus a SQL analytics tool for aggregations and trends. Writes: create and manage keywords and feeds — including wiring up Slack, email, and webhook notification destinations.
v2 is the default for everyone signing in to or signing up for Octolens. If you’re still on the v1 endpoint, see the migration note below.

​
Authentication

v2 uses OAuth — there’s no API key or token to paste. On first use, your MCP client opens a browser tab, you sign in to Octolens and approve the connection, and you’re ready. Your token never lives in a URL or config file. This is a change from v1, which embedded an API token in the connection URL.

​
Connection URL

Add this URL in your AI client and complete the OAuth sign-in. Step-by-step instructions for each tool are in the Setup Guides.
MCP is intended for interactive AI-client workflows. For scripts, backend jobs, or production integrations, use the REST API.

​
Setup Guides

Most popular AI tools support connecting to the remote MCP server directly. Follow our step-by-step guides: You can also generate setup snippets from inside the app under Settings → MCP, and try the server live in the Playground.

​
Tools

v2 ships 23 tools across mentions, keywords, feeds, tags, suggestions, workspace, usage, and analytics. See the full Tools Reference for what each one does and when your assistant will reach for it. For ready-to-paste prompts and workflows, see Use Cases.

​
Migrating from v1

The v1 MCP server is still available for existing users but is no longer the default. To move to v2:
  1. Remove the old v1 connection (the URL containing ?token=...).
  2. Add the new v2 URL https://app.octolens.com/api/mcp/v2 following the Setup Guides.
  3. Complete the OAuth sign-in when prompted.
v2 covers everything v1 did, plus feeds, analytics, keyword management, and notification destinations. See the v1 docs if you need to reference the legacy setup.

​
Security: mentions are untrusted content

Every mention Octolens delivers was written by an unknown third party somewhere on the internet. When your AI assistant reads mentions through this MCP server, that content is data, not instructions — a post could contain text deliberately crafted to steer an AI agent (“prompt injection”). Octolens scores mentions for relevance and sentiment, but does not remove or neutralize adversarial text. Practical guidance:
  • Be deliberate about combining Octolens with write-capable tools (email, code, payments, file access) in the same agent session. Prefer read-only sessions when browsing mentions.
  • If your agent acts on mentions automatically, gate consequential actions behind a confirmation step.
  • Tell your agent explicitly that mention text is quoted third-party content and never a source of instructions.

​
Terms of Use

The Octolens MCP server is intended for use with MCP-compatible AI clients (such as ChatGPT, Claude, and Cursor) for interactive, AI-driven workflows. Not permitted: Using the MCP endpoint via custom scripts, bridges, or standalone HTTP/JSON-RPC clients is unsupported and not permitted. The MCP server is not a general-purpose programmatic API. For programmatic access: Use the Octolens API instead — see the API documentation.