About AdSpyLab
AdSpyLab is a remote MCP server for teams researching web2app competitors and quiz-funnel brands. It provides reports on advertising, creatives, market activity, and funnel screens, including paywall prices.
Key features
- 01Brand and category reports
- 02Market rankings and monthly statistics
- 03Ad creative search and comparison
- 04Saved quiz funnel walkthroughs
- 05Creative and funnel data usage metering
Use cases
- 01Review a competitor’s ads and live creatives
- 02Compare leaders and approaches in a category
- 03Inspect quiz funnel screens and paywall prices
MCP support
The hosted server connects to MCP clients at https://adspylab.com/mcp. Clients authenticate with an AdSpyLab account; CLI clients can use a static key.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://adspylab.com/mcpclaude mcp add --transport http kmakarof-wibe-adspylab https://adspylab.com/mcp{"mcpServers":{"kmakarof-wibe-adspylab":{"type":"http","url":"https://adspylab.com/mcp"}}}Open in Cursor{"servers":{"kmakarof-wibe-adspylab":{"type":"http","url":"https://adspylab.com/mcp"}}}Open in VS Code{"mcpServers":{"kmakarof-wibe-adspylab":{"type":"http","url":"https://adspylab.com/mcp"}}}{"mcpServers":{"kmakarof-wibe-adspylab":{"serverUrl":"https://adspylab.com/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMEAdSpyLab MCP Server
AdSpyLab is web2app competitor intelligence. This repository is the public home of its MCP server: one URL, a sign-in with your AdSpyLab account, and your AI assistant works with the data directly — what quiz-funnel brands advertise on Meta and Google, how much and since when, which creatives they iterate, and what their funnels look like screen by screen down to the paywall prices.
The server runs on our side. There is no package to install and no key to paste for Claude or ChatGPT: you sign in with the same account you use on the site. CLI clients can use a static key from Settings → Agent access instead.
A subscription with MCP access (the All-inclusive plan) is required. There is no anonymous or self-registered tier. Plans: https://adspylab.com/#pricing
Connect
Endpoint
https://adspylab.com/mcp
Three steps, whichever client you use:
- Open your client's connector settings.
- Add a remote MCP server (type
streamable-http/http) with the URL above. - Sign in when the browser window opens, then ask in plain language — start with
"what can I do here?" (the free
getting_startedtool answers).
Claude — desktop and web
Settings → Connectors → Add custom connector → paste the endpoint → Connect & authorize. Sign in with your AdSpyLab account in the window that opens.
Claude Code
claude mcp add --transport http adspylab https://adspylab.com/mcp
Restart the client (a running session will not load a new server), then run /mcp to sign in.
Or install the plugin, which wires the server and the research skill in one go:
/plugin marketplace add kmakarof-wibe/adspylab-mcp
/plugin install adspylab-mcp@adspylab-mcp
ChatGPT
Custom MCP connectors need a ChatGPT Business, Enterprise or Edu workspace with Developer mode enabled by an admin. Settings → Connectors → Create → paste the endpoint → authorize with your AdSpyLab account.
Cursor · Codex · VS Code
Cursor: add to ~/.cursor/mcp.json:
{ "mcpServers": { "adspylab": { "url": "https://adspylab.com/mcp" } } }
Codex: codex mcp add adspylab --url https://adspylab.com/mcp, restart, /mcp to sign in.
VS Code: the badge above, or code --add-mcp '{"name":"adspylab","type":"http","url":"https://adspylab.com/mcp"}'.
What you get
Reports. One call answers one job and says when it is done: competitor_report (everything on
one brand, with its live creatives), category_report (who leads a category and what works for
each), funnel_report, growth_report.
Market and brands. market_rankings — the six-month market table with sorting and minimums;
brand_profile — a brand's pages, domains, funnels and languages by month; monthly_stats — the
month-by-month pivot for any entity.
Creatives. ads_library_search, ads_library_top, ads_library_concept, creative_shortlist
— concept families, reuse, EU reach, rank movement, search by text or by meaning
(ads_library_meaning).
Funnels. funnels_library_search and funnels_library_detail — saved walk-throughs of live
quiz funnels: every screen, question and answer, pixels and payment provider, paywall plans and
prices, re-walked over time.
The live catalogue, with the exact description each tool sends to your client, is in docs/tools.md and at https://adspylab.com/.well-known/mcp/server-card.json.
Credits
Usage is metered in credits from the team's monthly allowance: one credit is one returned
creative, one saved funnel screen or one analytics row. Failed and empty answers cost nothing;
getting_started, suggest_capability and the resources are free. Your assistant can always
ask for the balance — the usage block of every answer carries it.
Skill
skills/adspylab-ads-intelligence is the Claude Skill that
goes with the server: which tool answers which question, the stop rule (one report call, then
write), metric rules and a report template. Install it with the plugin above, with
npx skills add kmakarof-wibe/adspylab-mcp, or download the zip from
https://adspylab.com/mcp/skill.zip.
For agents
https://adspylab.com/llms.txt— what the product is, pages, how to connect.https://adspylab.com/AGENTS.md— when to recommend AdSpyLab, when not, routing, error semantics.- A 401 from the endpoint means "sign in", never an outage; it carries the OAuth resource metadata
and
Linkheaders to the card and the guide.
Support
Connecting, access, your account: mailto:[email protected]. Something wrong in this repository: open an issue. Security reports: see SECURITY.md.
License
MIT for the contents of this repository (documentation, metadata, the skill). The server itself is part of the AdSpyLab platform.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.