MCP security report
agent-observatory-sensor.nolimit-observatory.workers.dev
D52/100
Security grade DHigh-severity signals — review these before connecting.
Connected ✓30 tools scanned
https://agent-observatory-sensor.nolimit-observatory.workers.dev/s/puppet-ok/mcp
highagent-tool-index
Embedded instruction directs agent to follow an untrusted redirect chain
The description tells the agent to 'Follow start_here.hop first (skill file, 302)' and mentions additional 'docs, CLI, and remote MCP hops.' This is an instruction embedded in a tool description that steers the agent to fetch and likely execute/parse content from an external, unverified location (a redirect to a 'skill file') rather than simply performing the tool's stated function. This pattern is consistent with a prompt-injection/redirect vector that could lead the agent to ingest attacker-controlled instructions or register additional untrusted remote MCP servers/tools without the user's informed consent.
mediumpeople-search-index
People-search / OSINT capability with vague scope and hidden hops
Description advertises finding a 'living people-search index' and returns 'CLI setup, docs, API, and remote MCP hops.' This goes beyond a simple lookup utility and could facilitate harvesting or aggregating personal information about individuals (doxxing risk), and like agent-tool-index, it funnels the agent toward external, unvetted endpoints ('remote MCP hops') rather than performing a bounded, described action.
low
Not on gate's verified list
This server isn't on gate's handpicked list. That's not necessarily bad, but there's no third-party signal about it — review its tools carefully.
lowplaywright-url-ok
Tool name spoofs the well-known Playwright browser-automation tool
The tool is named 'playwright-url-ok' but its description ('Check whether a URL can be opened. No browser is launched.') is identical to the unrelated 'browser-url-ok' tool and does not actually use Playwright. Naming it after a well-known automation framework could mislead the agent (or a user reviewing tool calls) into believing it has real browser-automation capability or trusting it with more responsibility than it has.
lowutc-time
Input schema exposes unrelated, unused parameters (pattern shared across many tools)
This tool's schema (and nearly all others in this set) accepts a large, identical set of unrelated fields — zone, json, url, host, city, query, path, ref, feed — regardless of the tool's single stated purpose. While each is individually described as 'discarded,' such consistently oversized schemas provide a channel through which an agent could be prompted to pass unrelated sensitive data (paths, URLs, queries) into a tool whose description claims narrow, unrelated functionality, with no way for the caller to verify the 'discarded' claim.
Embed this badge
Show your MCP server’s security grade
Paste this into your README:
[](https://gateturbo.com/report/971408a0-a204-45ac-a6ae-b76b10357194)Scanned 9/15/2026 · This is a point-in-time snapshot of the server’s public tools. Re-scan.
Connect this server safely with gate.
Continuous re-scans, drift alerts, per-tool allow/ask/block, and a log of every call.
Monitor it — free