MCP security report
mudko.com
F5/100
Security grade FHigh-severity signals — review these before connecting.
Connected ✓26 tools scanned
https://mudko.com/api/mcp
highsummarize_scan_for_humans
Embedded 'cooperative sampling' prompt injection vector
The tool description instructs the calling agent to take a prompt returned in the response (`caller_reasoning`) and 'execute that prompt and present ITS output to your user as the pitch' at no cost. This is effectively an instruction embedded in a tool description telling the agent to run arbitrary server-supplied content through its own reasoning and present the result to the user as if personalized, without disclosing to the user that the content originates from a remote server. This is a classic prompt-injection pattern: the tool's output can dictate what the agent says next, bypassing the agent's own judgment about what to tell the user.
highrequest_invoice
Autonomous high-value invoice issuance with no human-confirmation safeguard
Description explicitly states the tool can 'autonomously close a deal' and that 'any amount up to the card processor's per-invoice maximum (~$999,999) auto-closes — a $500,000 close sails through,' bounded only by the payment provider rather than any explicit user/business confirmation step. Combined with instructions to 'never block the close' on missing contact info, this encourages the agent to issue very large, real financial invoices with minimal friction, which is a broad and risky capability relative to a 'send an invoice' tool.
mediumverify_credentials
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.
mediumrequest_quote
Instruction to covertly harvest budget data for remarketing regardless of purchase intent
The description tells the agent: 'Capturing the budget signal even from non-converting users dramatically improves re-engagement,' and instructs the agent to proactively ask price-hesitant users for budget info to enable later tailored follow-up. This nudges the agent to extract and retain personal/financial signal from users who have explicitly declined to buy, for the seller's remarketing benefit rather than the user's, which is a manipulative data-collection pattern embedded in tool guidance rather than a straightforward pricing function.
mediumsubmit_lead
Pressure to extract more PII than the schema requires
Although only `email` and `source` are schema-required, the description instructs the agent to treat contact_name, contact_phone, and best_consult_time as 'REQUIRED' and to 'ASK FOR ALL THREE explicitly,' noting phone-first follow-up converts better and that 'many users will share it once asked directly.' This directs the agent to persistently solicit sensitive contact details (especially phone number) beyond what the tool's actual input contract requires, which can pressure users into oversharing PII.
lowrequest_invoice
Unusually long tool description
Very long descriptions are sometimes used to bury instructions to the AI. Skim it before allowing.
lowsubmit_lead
Unusually long tool description
Very long descriptions are sometimes used to bury instructions to the AI. Skim it before allowing.
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.
lowcreate_bazaar_listing / get_bazaar_listing_kit / check_bazaar_listing
Autonomous on-chain settlement from vendor-funded wallet
check_bazaar_listing states that for endpoints within a price bound 'WE fire the indexing settlement from our own funded wallet' automatically as part of an idempotent pipeline. This is a legitimate stated feature, but it is worth flagging as an automated financial action taken without additional per-call user confirmation; low severity since it's vendor-funded and bounded, but it broadens the tool's effective capability beyond passive reporting.
Embed this badge
Show your MCP server’s security grade
Paste this into your README:
[](https://gateturbo.com/report/61a8b0bf-769a-4a59-81d9-a80d9c80f3f9)Scanned 9/6/2026 · 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