OtterlyAI Public API and Claude Skill: Automate AI Search Monitoring

What is the OtterlyAI Public API?

The OtterlyAI API gives you programmatic access to brand visibility data from the OtterlyAI platform — including brand KPIs, prompt-level coverage, citation sources, recommendations, and on-demand GEO Audits across six major AI search engines: ChatGPT, Google AI Mode, Google AI Overviews, Gemini, Perplexity, and Copilot.

Unlike a typical analytics export, the API is built for live workflows: it powers the OtterlyAI Claude Skill, an MCP server (coming soon) for Claude Code and Cursor, and any custom integration you want to wire into Looker, Slack, Google Sheets, or your CMS.

The base URL is https://data.otterly.ai, all endpoints are versioned under /v1, and authentication is a Bearer token. The full machine-readable spec lives at data.otterly.ai/v1/openapi.json.

Why this launch matters for GEO teams

Generative Engine Optimization is no longer a quarterly audit. According to OtterlyAI’s 2026 GEO research, Google AI Overviews now serves roughly 15 billion AI-assisted queries per day and ChatGPT another 2.5 billion, while 15% of total website traffic comes from AI agents and bots. The teams winning in AI search are the ones who can ship a competitor delta on Friday using Monday’s data — without filing a ticket with a data engineer.

The Public API and Claude Skill close that loop. Your dashboard becomes a data source, and Claude becomes the analyst sitting next to it.

The three components of OtterlyAI’s API

All three are backed by the same API key and the same workspace permissions.

  1. Public API (available today): Direct HTTP access to every report, prompt, citation, recommendation, and audit. Best for people building custom dashboards, BI pipelines, or webhook automations.
  2. Claude Skill (available today): A drop-in .skill file for Claude Desktop. Install once, then ask Claude in plain English. Best for marketers and GEO practitioners who don’t want to think about endpoints.
  3. MCP Server (coming soon): A hosted Model Context Protocol server at data.otterly.ai/mcp that exposes 11 read-only tools to any MCP-compatible client. Works with Claude Code, Claude Desktop, and Cursor.

9 GEO workflows you can automate today

The fastest way to understand the launch is to steal the workflows we run internally.

1. Weekly Brand KPIs for the CMO

Open Claude and ask:

Use the OtterlyAI skill to pull brand coverage for our brand vs. our top 5 competitors in brand report (id: XXX) across all tracked AI models for the last 7 days. Break it down by model. Write a 150-word executive summary, then post it to Slack #gtm-content-marketing.

2. A live, branded AI search dashboard you can re-open

Hand Claude a brand report ID, a country, and a date range. It pulls reports, prompts, citations, and stats, then renders a single self-contained, OtterlyAI-styled HTML dashboard with:

3. Close the GSC ↔ AI search blind spot

Plug Google Search Console data alongside the API. Claude surfaces the top 20–30 question-style queries you already rank for on Google and classifies intent, providing high leverage GEO opportunities.

4. Quarterly account audit in a board-ready PDF

Point Claude at a brand report and ask it to audit the setup: are you tracking the right prompts, how do you stack up against your top 5 competitors, and what are the three strategic priorities for next quarter?

5. Content briefs that target your citation gaps

Ask Claude to find the prompts where your brand and domain coverage are lowest, then pull citation reports to create a complete content brief.

6. Analyze Topic Clusters & Product Associations

Use the otterly skill to extract and analyze how AI engines describe your brand and competitors for actionable insights.

7. Recommendations -> Content Brief

Show my create_content_on_page and improve_existing_content recommendations for the US.

8. Recommendations -> Reddit Community Brief

List Reddit thread opportunities for our brand ranked by potential impact.

9. Run a content audit + citation gap analysis

Run a content audit in the OtterlyAI workspace for the following URLs attached.

How to set up OtterlyAI in Claude (in under 5 minutes)

Step 1: Generate an API key

  1. Sign in to the OtterlyAI dashboard.
  2. Go to API Keys.
  3. Click Create API key, choose the workspaces and permissions, then copy the key.

Step 2: Install the Claude Skill

  1. Download otterly.skill.
  2. Open Claude Desktop → Customize → Skills → + → Create Skill → Upload Skill.

Step 3: Tell Claude your API key

At the start of any session, paste:

My OtterlyAI API key is oai_live_xxxx

Optional: Hit the REST API directly

curl https://data.otterly.ai/v1/health \
  -H "Authorization: Bearer YOUR_API_KEY"

Available Today

The OtterlyAI API is available today for all Standard, Premium or Enterprise customers. If you are unsure about your OtterlyAI license, please get in touch via our live chat or at hello@otterly.ai. We are more than happy to help.

Frequently asked questions

Which AI engines does the OtterlyAI API cover?

The API exposes brand visibility data across six AI search engines: ChatGPT, Perplexity, Google AI Mode, Google AI Overviews, Gemini, and Copilot.

Is the OtterlyAI Public API free?

The API is included with the OtterlyAI Standard, Premium or Custom Enterprise plans. Generate a key from your dashboard.

What is the difference between the OtterlyAI MCP Server and the Claude Skill?

The Claude Skill is a packaged .skill file that wraps the same API into a natural-language interface specific for Claude Desktop. The MCP Server will be available soon.