PreClick — An MCP Native URL Preflight Scanning Service For Autonomous Agents (Formerly URLCheck).
Preflight URLs before your agent fetches or clicks them so phishing, mismatched intent, and risky redirects get caught early.
Overview
ai.urlcheck/urlcheck-mcp is a MCP server for the Ship phase that pre-scans URLs for threats and intent match before autonomous agents click or fetch them.
What is this MCP server?
- PreClick (URLCheck) scans links for threats and verifies intent match before agents click or fetch
- Hosted MCP at https://urlcheck.ai/mcp with optional X-API-Key header
- Trial tier: up to 100 requests per day without a key; API key for higher limits
- MCP-native design for autonomous agents and coding-agent browser workflows
- Server version 0.1.5 with source at cybrlab-ai/preclick-mcp
- Hosted trial up to 100 requests per day without API key
- Remote https://urlcheck.ai/mcp
Community signal: 6 GitHub stars.
What problem does it solve?
Coding agents that open user-supplied or retrieved links can walk into phishing or wrong-page actions with no built-in URL safety gate.
Who is it for?
Builders shipping browser-capable or fetch-heavy agents who want a low-friction URL preflight API via MCP.
Skip if: Pure offline CLIs with no URL access, or teams that already mandate a full corporate secure web gateway with no agent-level hook.
What do I get? / Deliverables
After wiring the MCP remote, your agent can reject or flag bad links before navigation, cutting incident risk in shipped agent workflows.
- Structured pre-click scan results for candidate URLs
- Threat and intent-mismatch signals usable in agent tool guards
- Trial-tier coverage without a key for early ship testing
Recommended MCP Servers
Journey fit
Autonomous browsing and tool-use explode once you ship agent features; URL safety belongs on the ship path alongside review and security hardening, not only in ad-hoc debugging. Security is the right shelf for threat scanning and intent-match checks on links before execution—core appsec for agent runtimes.
How it compares
MCP URL preflight scanner, not a general web scraping skill or host-level antivirus.
Common Questions / FAQ
Who is ai.urlcheck/urlcheck-mcp for?
It is for developers running Claude Code, Cursor, Codex, or custom MCP clients whose agents follow links and need pre-click threat and intent checks.
When should I use ai.urlcheck/urlcheck-mcp?
Use it before agent navigation, automated research fetches, or any workflow where untrusted URLs appear in prompts, search results, or tool outputs.
How do I add ai.urlcheck/urlcheck-mcp to my agent?
Add remote https://urlcheck.ai/mcp (streamable-http) to MCP config; set header X-API-Key when you need more than the hosted trial of 100 requests per day.