Catch brand impersonators from your AI agent.
The notolens MCP server gives your agent the same investigated matches, risk scores and source records as the dashboard. It sets up monitoring, triages the inbox and tracks self-guided takedowns. Every decision stays yours.
Included on every plan. No API keys: you sign in with your notolens account.
Triage from the conversation.
Ask what came in overnight. Your agent reads the inbox highest risk first, explains each match from its record, clears the noise and logs the takedown steps you take.
What came in overnight? Triage it for me.
- list_matches3 in inbox
One needs you today. notolens-login.com copies your sign-in page.
notolens-login.com
Sign-in form posting off-site
9/10
NOTOLENS PRO
EUIPO filing, class 25
6/10
notolnes.net
Parked with Sedo
2/10
Watch the filing, dismiss the parked one. Options on the phishing site?
- update_matches2 moved
- get_match3 guides
Done. Possible next steps: anti-phishing lists, registrar and hosting abuse reports, each with verified contacts and steps.
Filed the reports. Track it as a takedown.
- start_takedownstarted
- add_takedown_eventstep logged
Takedown started, reports logged. Mark it succeeded once the site is down.
Connect your client.
Add the server URL, then sign in with your notolens account when the client asks. Any client with remote MCP and OAuth support works the same way.
One-click install
Claude
Opens the add connector dialog with notolens filled in. Works on claude.ai, Claude Desktop and mobile.
Connect to ClaudeServer URLhttps://notolens.com/mcpCursor
Install in one click, or add the server to your MCP settings.
Add to Cursor~/.cursor/mcp.json{ "mcpServers": { "notolens": { "url": "https://notolens.com/mcp" } } }VS Code
Install in one click, or add the server to your MCP configuration.
Install in VS Code.vscode/mcp.json{ "servers": { "notolens": { "type": "http", "url": "https://notolens.com/mcp" } } }
Command line and other clients
Claude Code
Run this in a terminal, then /mcp in Claude Code to sign in.
Terminalclaude mcp add --transport http notolens https://notolens.com/mcpCodex CLI
Run this in a terminal, then codex mcp login notolens to sign in.
Terminalcodex mcp add notolens --url https://notolens.com/mcpGemini CLI
Run this in a terminal, then /mcp auth notolens in Gemini CLI to sign in.
Terminalgemini mcp add --transport http notolens https://notolens.com/mcpChatGPT
Add a custom MCP server with this URL in the ChatGPT connector settings, where your plan offers them.
Server URLhttps://notolens.com/mcpZed
Add the server to your Zed settings.
settings.json{ "context_servers": { "notolens": { "url": "https://notolens.com/mcp" } } }Other clients
Add a remote MCP server over Streamable HTTP with this URL and sign in when asked.
Server URLhttps://notolens.com/mcp
You decide what the agent can change.
The agent works inside your account. It reads what notolens found and changes only what you ask it to.
Sign in, no API keys
Your client opens notolens in the browser. You sign in, or create an account, and approve the connection.
Read-only or write
Uncheck the write permission on the consent screen and the agent can look but change nothing.
Your calls
The agent acts only through your account. notolens never dismisses a match or submits a takedown on its own.
Confirmed first
Adding an asset, previewing a site and sending feedback each need your explicit go-ahead.
Some steps stay in the dashboard
Signup, billing, archiving assets and deleting monitors happen on notolens.com.
Disconnect any time
Settings lists every connected app. Disconnecting one revokes its access right away.
21 tools, one inbox.
Setup, triage and takedown tracking from the conversation. Write tools need the write permission.
Overview
- Read
get_accountConfirms the signed-in account, its plan and limits, and whether the connection can write.
- Read
summarize_matchesCounts matches by status and kind, including how many are new.
- Read
list_scansShows when each monitor last checked its sources and what it found.
Monitors and assets
- Read
list_monitorsLists each brand monitor and the domains, marks and app names it watches.
- Write
create_monitorCreates a monitor for one brand. It starts empty until an asset is added.
- Write
update_monitorPauses or resumes a monitor. Its matches stay in place.
- Write
add_assetAdds a domain, mark or app name to a monitor after the user confirms the exact value.
- Write
preview_domainRenders the official site so the user can approve it as the visual baseline before adding.
- Write
prepare_asset_uploadStages a mark image, app icon or screenshot the user supplies.
Matches
- Read
list_matchesLists matches in one status, highest risk first, with risk and verdict filters.
- Read
get_matchReturns the risk score, plain-language summary, full source record and possible next steps.
- Read
get_match_screenshotReturns the captured screenshot of a domain match.
- Read
get_match_evidenceReturns a short-lived download link for the evidence PDF.
- Write
update_matchesMoves matches between inbox, watching, resolved and dismissed, by id or by filter.
Takedowns
- Read
list_takedownsLists open or closed takedowns with links back to their matches.
- Read
get_takedownReturns the full takedown timeline and the match it tracks.
- Write
start_takedownOpens a self-guided takedown on a match the user decides to act on.
- Write
update_takedownCloses a takedown as succeeded, unsuccessful or dropped, or reopens it.
- Write
add_takedown_eventLogs a comment or a step the user performed on the takedown timeline.
- Write
retract_takedown_eventMutes a mistaken comment or step. The entry stays on the timeline.
Feedback
- Write
send_feedbackEmails a question or bug report to notolens once the user approves the text.
Prompts
review_inboxWalk the user through their undecided matches, worst risk first, and triage them on their decisions.
Technical details.
For client developers and directories. The server card carries each tool's input and output schemas.
- Server URL
- https://notolens.com/mcp
- Transport
- Streamable HTTP
- Protocol versions
- 2025-11-25, 2025-06-18, 2025-03-26
- Authorization
- OAuth authorization code flow with PKCE (S256); clients identify through client ID metadata documents or dynamic client registration
- Scopes
- mcp:read, mcp:write (optional), offline_access
- Resource metadata
- https://notolens.com/.well-known/oauth-protected-resource/mcp
- Server card
- https://notolens.com/.well-known/mcp/server-card.json
- Registry name
- com.notolens/notolens
- Support
- Contact form
- Privacy
- Privacy policy
- Terms
- Terms of service
Questions about the MCP server
Something missing? Talk to us.
No. MCP access is included on every plan, Domain Monitor included. Signup and billing happen on notolens.com; once the account has a plan, the agent can do everything the plan allows.
notolens never dismisses a match on its own. Your agent acts only through your account, so what it triages is your call, and any match can be moved back to the inbox. Dismissed domain matches also return to the inbox when the domain reappears.
Start the sign-in again from the client: /mcp in Claude Code, codex mcp login notolens in Codex, /mcp auth notolens in Gemini CLI, or reconnect the connector in Claude. Sign-in needs a browser on the same machine, so it does not complete inside a remote SSH session.
The same matches, risk scores, plain-language summaries and possible next steps as the dashboard, plus the full observed record behind each match: site, mail, hosting and registration facts for domains, the register entry for trademarks, and the store listing for apps.
Brand protection that starts in minutes.
Add your domain and monitoring starts immediately. Every match arrives investigated, in the dashboard, your daily digest, or your AI agent.
Start monitoring in 3 minutes