All MCP Servers
OtherMCP RegistryremoteB · 68

Bannerproof MCP server

app.bannerproof · Updated

EU cookie-consent audit: trackers before consent, Reject all, Consent Mode v2, TCF; evidence + fixes

At a glance

Bannerproof is an MCP server in the Other category. It is hosted remotely and connects over HTTP, and declares no required credentials. It scores B (68/100) on the MCP Skills Hub health score as of Oct 28, 2026.

Source 0GitHub stars

One-click install

Opens your editor and pre-fills the config. Your editor will ask before adding it.

How to install Bannerproof in your AI client

Setup for Claude Code, Claude Desktop, Cursor, Codex, Windsurf, Cline and VS Code — copy-paste ready.

How to install Bannerproof in Claude CodeAnthropic's agentic CLI

  1. Add the server

    bash
    claude mcp add --transport http bannerproof https://api.bannerproof.app/mcp
  2. Verify it's connected

    bash
    claude mcp list
  3. Sharing it with your team?

    Add `--scope project` to write it to .mcp.json in the repository root, which you can commit (with the secrets left as environment variables).

  4. Use it in a session

    Ask Claude to use the tools this server exposes. It discovers them automatically.

What tools does Bannerproof expose?

This server requires credentials before it will list its tools, so we cannot enumerate them here. Connect it in your client to see what it offers.

What the code can do

This is a remote server — nothing is installed or run on your machine, so there is no local code to analyze. What it can reach is described above: your requests go to the provider's host.

Before you install

Signals gathered from the source repo and package registry — not a safety verdict.

What it can reach

  • Third-party hosted service — Your requests and any data in them are sent to api.bannerproof.app.
  • No local execution — Nothing is installed or run on your machine.
  • Last commit today — Actively maintained.
  • Repository created today — Very new project with little track record yet.
  • 0 GitHub stars — Very little community usage so far.

Client compatibility

Claude Code: NativeClaude Desktop: Via mcp-remoteCursor: NativeCodex CLI: Via mcp-remoteWindsurf: NativeCline: NativeVS Code: Native

Health score

B · 68

Maintenance, adoption, transparency and packaging — not a safety audit. How this is calculated

  • Maintenance30%100%

    Last commit 0 days ago.

  • Adoption20%0%

    0 stars.

  • Transparency20%75%

    3/4: ✓ public source repo, ✓ declared license, ✓ documentation, ✗ discoverable tool list.

  • Provenance15%75%

    Links to its own repository; package linkage unverified.

  • Packaging15%75%

    Remote endpoint; authentication requirements not declared.

Documentation

Full README

Bannerproof — cookie-consent audits for AI assistants and CI

Bannerproof checks a website the way a real first-time visitor from the EU sees it, and reports:

  • trackers, analytics/ad tags and tracking cookies that load before consent;
  • whether the cookie banner has “Reject all” on the first layer, and what still fires after Reject;
  • Google Consent Mode v2 state per signal (ad_storage, analytics_storage, ad_user_data, ad_personalization);
  • IAB TCF consent string before a choice and after Reject;
  • Global Privacy Control (informational in the EU);
  • cookies the site sets that its own cookie policy doesn’t mention;
  • third-party embeds that leak visitor data before consent (Google Fonts, YouTube, Maps, reCAPTCHA).

Every finding comes with the exact request as evidence, a concrete fix (13 languages) and the legal reference (ePrivacy Art. 5(3), GDPR, national law). Works with any consent platform — Cookiebot, Usercentrics, OneTrust, Complianz, CookieYes, Borlabs, iubenda or a custom banner.

This repository holds the public pieces: the MCP server listing, connection examples and a GitHub Action. The service itself runs at bannerproof.app.


1. MCP server (Claude, ChatGPT, Cursor, …)

Remote MCP server, Streamable HTTP, OAuth 2.0:

text
https://api.bannerproof.app/mcp

A free quick check works without an account (https://api.bannerproof.app/mcp/free). Full findings, monitoring, audits and evidence exports use a Bannerproof account — sign in with OAuth when your client asks, or use an API token from Settings.

| Client | How to connect | |---|---| | Claude (claude.ai / Desktop) | Settings → Connectors → find Bannerproof in the directory, or add a custom connector with the URL above | | Claude Code | claude mcp add --transport http bannerproof https://api.bannerproof.app/mcp | | Cursor | see `examples/cursor-mcp.json` | | Cline | see `llms-install.md` | | Other clients | add an MCP server with the URL above (OAuth), or send Authorization: Bearer <API token> |

Step-by-step guide: bannerproof.app/help/connect-claude-chatgpt

Tools

| Tool | What it does | Account | |---|---|---| | quick_check | yes/no + score for any public site | free | | explain_finding | what a finding means, why it matters, how to fix it, legal basis | free | | get_capabilities | what the server checks and what each tool does | free | | get_findings | full findings with evidence, fixes and legal references | yes | | scan_domain | fresh scan of a monitored domain | yes | | compare_scans | what changed since the previous scan | yes | | get_evidence | evidence pack (every request with phase and time, cookies, consent signals) as CSV/JSON | yes | | get_report_link | private report link to send to a client or developer | yes | | audit_domain / audit_domains | audit any site, or up to 10 at once (1 audit credit each; failed scans are refunded) | yes | | get_audits_status | status of several audits at once | yes | | list_domains, add_domain, scan_history, account_info, get_upgrade_link | account helpers | yes |

Maintain Bannerproof? Add a badge to your README

Shows its health grade, and links readers to install steps for every client.

markdown
[![Bannerproof on MCP Skills Hub](https://mcp-skills-hub.vercel.app/badge/app-bannerproof-bannerproof)](https://mcp-skills-hub.vercel.app/mcps/app-bannerproof-bannerproof)

Catalog data refreshed . Something wrong? Report it.