Skip to content
Official MCP RegistryListed

Prompt Sanitizer — prompt-injection scan for untrusted content

Part ofPrompt Sanitizer — prompt-injection scan for untrusted contentlisted on 2 directories

Scan web pages, mails, PDFs and tool output for prompt injection before your agent reads them.

First seen 14 Nov 2026. Evidence as of 17 Nov 2026.

4
Tools
From an anonymous probe
1
Source listings
Each with its own history
10
Recorded changes
Since first seen

Tools

ToolDescriptionBehaviour
buy_pack_with_cardFor agents and people without USDC: prepaid packs paid in EUR by card or Link (100 calls €1.50 · 500 calls €5.00 · 5,000 calls €39.00) — including Link's wallet for agents (one-time card, approved by the human). Returns checkout_url; after payment the success page (Accept: application/json) returns the bundle_token, and it is e-mailed. After: get_signals with bundle_token.Changes data
input_guardPrompt-injection and hidden-instruction scan for untrusted input before an agent acts on it: web pages, emails, RAG chunks, tool and MCP responses. Flags text aimed at the model (role takeover, forged system or tool turns, requests for keys, the prompt, history or payments, exfiltration links), invisible Unicode (tag smuggling decoded, zero-width, bidi), homoglyphs, base64 payloads and HTML hidden from humans, in 17 languages. Verdict clean, caution or hostile, score, findings, cleaned text. 0.015 USDC base (new payers 0.015, >10 payments 0.0105; 1 pack call). After: preflight before your agent pays anything the content pointed to (the result's next field has the exact call).Read-only
input_guard_previewFree prompt-injection quick scan for untrusted input: up to 4,000 characters of text, or one https page, checked for hidden instructions aimed at agents (role override, invisible Unicode, hidden HTML) in 17 languages incl. Chinese, Japanese, Korean, Hindi, Arabic. Returns verdict, score and the top 3 findings. After: input_guard for every finding and the cleaned text.Read-only
stealth_readRead a public https page for an agent: the main content as clean Markdown (headings, lists, tables kept; navigation, cookie banners, scripts and everything hidden from humans removed), scrubbed by the input guard — instructions aimed at the model replaced, invisible Unicode stripped — with verdict, score, findings, title and language. JSON and text pass through. One GET as a declared bot, no login or bot-protection bypass; a blocked page is never charged. 0.02 USDC base (new payers 0.02, >10 payments 0.014; 1 pack call). After: preflight before your agent buys or pays anything the page offers (the result's next field has the exact call).Read-only

Change history

  1. stealth_read: annotations changed
  2. input_guard_preview: annotations changed
  3. input_guard: annotations changed
  4. buy_pack_with_card: annotations changed
  5. server instructions changed (+"first call" +"agent free," +"per check")
  6. stealth_read: tool added
  7. input_guard: description changed (+"0.015" +"0.015," +"0.0105; 1")
  8. buy_pack_with_card: description changed (+"€1.50" +"€5.00" +"€39.00)")
  9. server instructions changed (+"0.015" -"0.99")
  10. Listed (registry)
Source listings
SourceListingFirst seenLast seenVersions
Official MCP Registrycom.oroboros-design.angel/prompt-sanitizer14 Nov 202617 Nov 20261
Prompt Sanitizer — prompt-injection scan for untrusted content on Official MCP Registry | InvokeRank