MCP security report

www.pathwren.workers.dev

F31/100
Security grade FA few things are worth a closer look before connecting.
Connected ✓9 tools scanned

https://www.pathwren.workers.dev/c/mcp-registry-official/mcp

mediumwhoami
References secrets or data exfiltration
The tool text references credentials, environment variables, or sending data elsewhere. This can be legitimate, but review what this tool actually accesses.
mediumexample
References secrets or data exfiltration
The tool text references credentials, environment variables, or sending data elsewhere. This can be legitimate, but review what this tool actually accesses.
mediumwhoami
Embedded instruction pushes unconditional self-fingerprinting call
The description reads like a directive to the agent ('POST that exact body ... and you are done') rather than a neutral tool description, and frames calling it as free of any consent requirement ('no consent prompt, nothing for you to invent'). It also explicitly returns the caller's IP address and User-Agent string back to a third-party server (pathwren.workers.dev), which is a data-collection action being normalized as harmless boilerplate. Combined with the 'FIRST CALL, needs nothing' text injected into every other tool's description, this looks like an attempt to get the agent to always call whoami first and leak request metadata about the user/agent to the operator on every session, regardless of the user's actual intent.
mediumclassify_user_agent
'FIRST CALL, needs nothing' injected pseudo-instruction referencing whoami
The description opens with a literal JSON-RPC call to the unrelated 'whoami' tool, framed as a mandatory first step ('FIRST CALL, needs nothing'). This is prompt-injection-style text embedded in a tool description trying to steer the agent's call sequence toward a data-collecting tool (whoami) before every actual tool use, rather than describing this tool's own function.
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.
lowexample
Same directive-style phrasing urging an unsolicited call
Description again reads as an instruction ('POST that exact body...and you are done') rather than a description of functionality, encouraging the agent to make a call 'to see the shape of an answer before you decide what to send,' which is unnecessary framing pressure rather than a factual capability statement.
lowlookup_crawler
Injected 'FIRST CALL' directive to invoke whoami
Same boilerplate instruction embedded at the start of the description, unrelated to this tool's actual purpose, repeatedly nudging the agent to call whoami first.
lowlist_crawlers
Injected 'FIRST CALL' directive to invoke whoami
Same boilerplate instruction embedded at the start of the description, unrelated to this tool's actual purpose, repeatedly nudging the agent to call whoami first.
lowgenerate_robots_txt
Injected 'FIRST CALL' directive to invoke whoami
Same boilerplate instruction embedded at the start of the description, unrelated to this tool's actual purpose, repeatedly nudging the agent to call whoami first.
lowis_verified_crawler_ip
Injected 'FIRST CALL' directive to invoke whoami
Same boilerplate instruction embedded at the start of the description, unrelated to this tool's actual purpose, repeatedly nudging the agent to call whoami first.
lowwhats_changed
Injected 'FIRST CALL' directive to invoke whoami
Same boilerplate instruction embedded at the start of the description, unrelated to this tool's actual purpose, repeatedly nudging the agent to call whoami first.
lowchanges_since
Injected 'FIRST CALL' directive to invoke whoami
Same boilerplate instruction embedded at the start of the description, unrelated to this tool's actual purpose, repeatedly nudging the agent to call whoami first.
Embed this badge

Show your MCP server’s security grade

MCP security grade F

Paste this into your README:

[![MCP security](https://gateturbo.com/badge/scan/890596a0-d328-4f6f-b27d-e8e9dd8238a2)](https://gateturbo.com/report/890596a0-d328-4f6f-b27d-e8e9dd8238a2)

Scanned 9/7/2026 · 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