A dashboard is a great place to look at your AI visibility. It’s a poor place to keep it locked. Today BotScope gets two ways out: a read-only REST API and a hosted MCP server — so your data can flow into your own tools, your reports, and the AI assistants you already work in.
TL;DR
- A read-only REST API at
api.botscope.ai/api/v1— projects, scans, L1–L4 scores, share of voice, competitors, categories and raw responses, as clean JSON. - A hosted MCP server at
mcp.botscope.ai— connect Claude (or any MCP client) and it can query your visibility data as tools, in plain conversation. - One credential for both — an API key you create in Settings → Developers, sent as a Bearer token. Same key, both surfaces.
- Included on every paid plan (Starter and up). Reads are free and rate-limited; they never touch your scan credits.
- Available now, with an OpenAPI spec and full docs.
Why we built it
The teams getting the most out of BotScope kept asking the same thing: can I get this data out? An agency wants visibility scores next to the rest of a client dashboard in Looker. An in-house team wants last night’s L4 citation count posted to Slack. A founder just wants to ask, in the middle of a conversation, “are we losing ground to that competitor in AI search?” — without leaving the chat to go read a chart.
All three are the same need: your data, where you already work. So we built the two surfaces that cover it — a conventional API for systems, and MCP for assistants.
The REST API
Plain HTTPS, JSON in a { "data": … } envelope, one Bearer key. Nine read endpoints cover the product:
export BSK="bsk_your_key_here"
export BASE="https://api.botscope.ai/api/v1"
# List projects, then read one project's headline visibility
curl -s -H "Authorization: Bearer $BSK" "$BASE/projects"
curl -s -H "Authorization: Bearer $BSK" "$BASE/projects/$PID/visibility"You get the same numbers the dashboard shows — overall and per-layer scores with trends, share of voice, competitor rankings, category landscape, and the raw model responses with their brand detections and citations. Filter by provider, pin a specific scan, or aggregate across a date window. Every response is org-scoped: a key only ever sees its own organisation’s data. Full reference and an OpenAPI 3.1 spec live in the docs.
The MCP server
MCP — the Model Context Protocol — is the emerging standard for giving AI assistants tools. Our hosted MCP server wraps the same API, so an assistant like Claude can call BotScope instead of you copy-pasting numbers into a prompt.
Connect it once, then just ask:
How did Acme’s AI visibility do this month, and how do we compare to competitors?
Acme’s overall AI-visibility score is 72, up +6.4 on last scan — its strongest month yet. The shape underneath:
- L1 Entity 100 — every model knows Acme exists.
- L2 Depth 88 — rich, accurate descriptions (up 4).
- L3 Recommendations 64 — named in most “best in category” lists (up 9).
- L4 Citations 41 — more of your pages are cited across providers (up 7).
On share of voice you’re leading at 58% across tracked queries — ahead of Globex (22%) and Initech (11%), and the gap widened this month. You’re winning at L1/L2 and now pulling ahead on recommendations; the main upside left is L4 citations on Perplexity, where Globex still edges you. Want me to pull the prompts where a competitor gets cited and you don’t, so you can target them?
Behind that answer, Claude called get_visibility and get_competitors, read the JSON, and wrote the summary. You didn’t open the dashboard — you asked a question and got an answer grounded in your real scan data. The same works in any MCP-capable client: Claude Code, Cursor, Windsurf, and more.
How to connect
You’ll need a key first: app.botscope.ai → Settings → Developers → Create key. It’s shown once — copy it somewhere safe.
MCP — Claude Code (one line, real key):
claude mcp add --transport http botscope https://mcp.botscope.ai/mcp --header "Authorization: Bearer bsk_your_key_here"MCP — Cursor / Windsurf / generic mcp.json:
{
"mcpServers": {
"botscope": {
"url": "https://mcp.botscope.ai/mcp",
"headers": { "Authorization": "Bearer bsk_your_key_here" }
}
}
}That’s it — your client will list nine BotScope tools and you can start asking.
A few things you can build
- Pipe scores into Looker / Sheets / your data warehouse on a schedule, next to the rest of your reporting.
- Slack a daily digest — overall score, biggest mover, any competitor that jumped.
- Let clients (or your team) self-serve — “ask Claude how we’re doing in AI search” beats another login.
- Trigger workflows when a layer drops or a competitor overtakes you.
FAQs
Is this included in my plan? Yes — the API and MCP server are available on every paid plan (Starter and up). API reads are free and rate-limited; they don’t consume scan credits.
What’s the difference between the API and the MCP server? Same data, two front doors. The REST API is for systems and scripts. The MCP server is for AI assistants — it exposes the API as tools they can call in conversation. Both use the same key.
How does authentication work? You create an API key in Settings → Developers and send it as a Bearer token. Keys are scoped to your organisation, shown once at creation, and revocable at any time.
Can it write or trigger scans? Not yet — both surfaces are read-only today. Your data is safe to expose to an assistant; it can read, not change.
Does it work with Claude’s one-click connectors? Today it connects via API key in clients that support custom headers (Claude Code, Cursor, Windsurf, and others). One-click OAuth connectors for Claude.ai are on the roadmap.
Where are the full docs? At botscope.ai/docs/api, including every endpoint, parameters, pagination, errors, and the machine-readable OpenAPI spec.
Your visibility data was never meant to live behind one login. Now it doesn’t have to — query it from your stack, your reports, or just by asking. We’re excited to see what you wire it into.