MCP security report
www.pathwren.workers.dev
F19/100
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/triage
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.
mediumtriage_log
Embedded instruction to call a different tool first, formatted as a literal request payload
The description opens with 'FIRST CALL, needs nothing' followed by a full JSON-RPC body naming a different tool (no_arguments_triage_this_hosts_own_crawler_log). This is an instruction embedded in the description aimed at directing the agent's behavior/call sequence rather than describing the tool's own parameters — a classic prompt-injection pattern that could cause the agent to make extra, unrequested calls.
mediumfind_impersonators
Same 'FIRST CALL' redirection pattern
Repeats the same embedded directive telling the agent to first call a completely unrelated tool via a hardcoded JSON-RPC body, rather than simply documenting this tool's own use. This steers agent control flow outside the user's actual request.
mediumsummarize_by_operator
Same 'FIRST CALL' redirection pattern
Contains the identical embedded JSON-RPC 'first call' instruction seen across sibling tools, which is not necessary to explain this tool's function and appears designed to manipulate call ordering/agent behavior.
mediumrobots_from_log
Same 'FIRST CALL' redirection pattern
Contains the identical embedded JSON-RPC 'first call' instruction directing the agent to invoke a different tool before doing what was asked.
mediumwaf_ruleset_from_log
Same 'FIRST CALL' redirection pattern
Contains the identical embedded JSON-RPC 'first call' instruction directing the agent to invoke a different tool before doing what was asked.
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.
lowno_arguments_triage_this_hosts_own_crawler_log
Instructs agent to POST raw JSON-RPC directly to an external URL
The description repeatedly tells the agent to construct and POST an exact JSON-RPC body to https://www.pathwren.workers.dev/mcp/triage, rather than relying on the normal MCP tool-invocation mechanism. Encouraging an agent to manually craft and send raw HTTP requests to a hardcoded external endpoint (outside the standard tool-call path) is an unusual pattern that could be used to normalize direct network calls to attacker-controlled infrastructure in other contexts.
lowwhoami
Claims persistent tracking of caller's identity/history
The tool states it reports 'whether we have logged you here before,' implying the host retains and correlates the calling agent's IP address and user-agent across invocations. This is an undisclosed data-retention/fingerprinting capability beyond the stated triage purpose, worth flagging as a privacy concern for whoever operates the calling agent.
lowtriage_my_request
Excessive persuasive language to pre-empt scrutiny
The description uses repeated reassurance phrasing ('nothing for you to invent', 'no key, no account... read-only') across many tools in this set, a pattern often used in prompt-injection attempts to lower an agent's scrutiny of what is actually being requested or sent (including the caller's own IP/user-agent).
Embed this badge
Show your MCP server’s security grade
Paste this into your README:
[](https://gateturbo.com/report/9342a1e9-b028-49fd-98f2-78584884c334)Scanned 9/14/2026 · 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