Skip to main content
Glama

Prompt Sanitizer — prompt-injection scan for untrusted content

Server Details

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

If you are the author of this connector, you can claim ownership. Claimed connector authors can inspect health checks, view analytics, and manage their listing.
Status
Healthy
Uptime
100.0% over the last 24 hours
Last Tested
Transport
Streamable HTTP · MCP 2025-11-25
URL
Repository
daoudandiaye98-alt/mcp-prompt-sanitizer
GitHub Stars
0
Server Listing
prompt-sanitizer
ChatGPT
–
Claude
–

TDQS

A4.1/5.0

Scored across 4 tools

Disambiguation4/5

The four tools split into reasonably distinct roles: a purchase flow (buy_pack_with_card), a full text scanner (input_guard), a free tier of that scanner (input_guard_preview), and a fetch-and-sanitize tool (stealth_read). The preview/full pair and the stealth_read-vs-input_guard split do overlap, but the descriptions state the boundaries explicitly (4,000-char free preview with top-3 findings vs. full findings plus cleaned text; URL fetch vs. caller-supplied text).

Naming Consistency3/5

All names use snake_case, which keeps casing consistent, and input_guard/input_guard_preview form a clear family. However the conventions are mixed: buy_pack_with_card is verb_object, input_guard is a bare noun phrase, and stealth_read is adverb+verb, so there is no single predictable pattern an agent can extrapolate from.

Tool Count4/5

Four tools is a tight, well-scoped surface for a scanning service, and each tool maps to a real workflow step (try free, scan, fetch-and-scan, pay). The only slight mismatch is that one of the four slots is consumed by a billing tool rather than by scanning capability, but it is necessary for the paid model.

Completeness3/5

Core scanning is covered for text, URLs, and (per description) tool/MCP responses, but the lifecycle has holes: there is no tool to check remaining quota or key status, no way to verify or poll a purchase after receiving checkout_url (the key arrives by e-mail or a success page), and no batch/multi-item scan. These are workaroundable but will cause dead ends for an agent operating autonomously on a credit budget.

Available Tools

4 tools
buy_pack_with_cardAngel: 5,000 calls by cardAInspect

For businesses without USDC: 5,000 calls €9.00, one-time, 90 days, businesses only, paid in EUR by card or Link. Returns checkout_url for a person to open; the checkout asks for the company or freelance name. After payment the success page (Accept: application/json) returns the key, and it is e-mailed. After: any tool with bundle_token.

Inputs: calls, email.

input_guardRemove hidden content from untrusted text and flag text aimed at the modelA
Read-only
Inspect

Removes hidden content and flags text aimed at the model, for untrusted input before an agent acts on it: web pages, emails, RAG chunks, tool and MCP responses. Strips invisible Unicode (tag smuggling decoded, zero-width, bidi) and HTML hidden from people; flags role takeover, forged system or tool turns, requests for keys, the prompt or payments, exfiltration links, homoglyphs and base64 payloads, in 17 languages. Rule-based. Verdict clean, caution or hostile, score, findings, cleaned text. Free with a key (500 calls) or 5 calls a day without one; then 0.02 USDC per call via x402, or 1 call from a key. After: preflight before your agent pays anything the content pointed to (the result's next field has the exact call).

Inputs: url, text, payment_tx, bundle_token, +1 more.

input_guard_previewPrompt-injection quick scan (free)A
Read-only
Inspect

Free 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.

Inputs: url, text.

stealth_readRead any public page as clean Markdown, with hidden text removedA
Read-only
Inspect

Returns a public https page as clean Markdown for an agent (headings, lists, tables kept; navigation, cookie banners, scripts and everything hidden from people removed), strips invisible Unicode and flags text aimed at the model — 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. Free with a key (500 calls) or 5 calls a day without one; then 0.02 USDC per call via x402, or 1 call from a key. After: preflight before your agent buys or pays anything the page offers (the result's next field has the exact call).

Inputs: url (required), payment_tx, bundle_token, payer_address.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      Not graded
      maintenance
      Enables agents and applications to detect, inspect, and neutralize prompt injection, jailbreak attempts, and hidden instruction escapes in untrusted text before it is passed into prompts.
      1
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables users to scan untrusted text, files, and directory trees for prompt-injection patterns and receive severity-ranked, explainable verdicts before content reaches an agent. It runs locally and returns clean, review, or blocked verdicts for safer agent input.
      1
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Protects AI agents from prompt injection attacks, jailbreak attempts, and common web vulnerabilities by screening untrusted input through semantic LLM analysis and static pattern matching.
      71 npm
      2
      ISC
    Try in Browser

    Glama MCP Gateway

    Add one secure layer between your agents and this server.