Skip to content

KillBounce Data MCP server

app.killbounce.data/toolsdata.killbounce.appstreamable-httpv1.2.0Scanned

Pay-per-call agent tools: web, email, domains, SEC, files, code, images, translation. USDC (x402).

27 tools: 3 can delete or cancel things, 2 do something that can't be undone, 19 change things and 3 only read.

Tools: 3 delete, 2 irreversible, 19 write, 3 read.
  • Reads private data · 1
  • Reads text others write · 27
  • Can send data out · 11
  • 1 finding

Tools listed without sign-in. Remote posture

With Anlyon

3 tools here can delete or cancel things and 2 do something that can't be undone. Govern them.

  • Anlyon lists the tools and adopts none.
  • Each tool you adopt gets approvals and receipts.
  • Your agent never holds the key.
Govern this server →

A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.

What the KillBounce Data MCP server can do

The KillBounce Data MCP server is published by data.killbounce.app, a domain the MCP Registry verified. It has 27 tools. It can delete or cancel things (3 tools: kv, file and code), do things that cannot be undone, such as sending, paying or publishing (email_send and translate) and create or change data (19 tools: email_verify, site_check, dns, markdown and 15 more). 3 tools only read.

1 tool reads private data (github_tree); 27 tools bring in text other people can write (email_verify, site_check, dns and 24 more); 11 tools can send data out of your system (markdown, meta, diff and 8 more). On its own it has all three parts of a prompt-injection exfiltration path: private data, text others write, and a way out.

The fixed checks matched 1 finding (annotation-conflict). Each one comes with the text it matched.

Its remote endpoint lists its tools without sign-in, and its tools changed once since Anlyon Radar first saw them.

Last scanned Nov 21, 2026, version 1.2.0.

Tools

Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.

27 tools

Deletes or cancels 3Claude Code: deny
  • codeKeep a code project (one or many files) with full version history, private to the paying wallet.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Keep a code project (one or many files) with full version history, private to the paying wallet. The original is never changed: every edit makes a new version (exact find-and-replace or a line range), returns a diff, and can branch from any earlier version. action=store $0.005 for projects up to 1 MB (then $0.005 per MB), kept 30 days, renewable. edit, get, diff, list and delete cost $0.0025. Store by POST JSON {files: {"path": "content"}}. Up to 50 MB, 500 files and 500 versions per project. An edit can also create a file (path + text) or remove one (remove=1). Price depends on the action and size, as above (USDC on Base, x402); you're only charged when the call succeeds.

    Annotations the server declares

    • title: Versioned code workspace (never lose work)
    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
  • fileStore files privately for the wallet (or card pack) that pays: only that payer can list or download them.Reads text others writeNo findings

    Why it is in this group

    • The description's first verb is "store": a word for a change.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Store files privately for the wallet (or card pack) that pays: only that payer can list or download them. action=store with bytes=<size> first returns the exact price (HTTP 402): $0.005 per MB, minimum $0.005, kept 30 days; action=renew extends 30 days at the same price. action=download costs $0.005 any size. list, info and delete cost $0.002. Upload by POST (raw bytes, or JSON {data_base64}); max 95 MB per file. Downloads come back as raw bytes, or add format=base64 for JSON (files up to 20 MB). Price depends on the action and size, as above (USDC on Base, x402); you're only charged when the call succeeds.

    Annotations the server declares

    • title: Private file storage for agents (long-term memory)
    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
  • kvSmall persistent memory that belongs to the wallet (or card pack) that pays: save, read, list and delete values between runs.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Small persistent memory that belongs to the wallet (or card pack) that pays: save, read, list and delete values between runs. Nobody else can read or overwrite your keys. Up to 1,000 keys of 10 KB each; values expire 90 days after their last write. action=set|get|delete|list (GET with query params, or POST JSON {key, value}). Price: $0.002 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Private key-value memory for agents
    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
Can't be undone 2Claude Code: ask
  • email_sendSend a plain-text email from your agent (from agents@mail.killbounce.app).Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "email_send" says "send": a word for sending, and what is sent cannot be taken back.
    • The input addresses people (input.to): what it sends cannot be taken back.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe name "email_send" says "send": it sends or publishes something to other people.
    • outboundThe input takes a destination (input.to): what the tool writes goes there.
    Description, as the server wrote it
    Send a plain-text email from your agent (from agents@mail.killbounce.app). The first call for a new address sends them a short request to allow your wallet (status permission_requested, $0.005), showing your subject line; the message itself is not delivered or queued, so call again after they accept. Every email names the sending wallet and carries stop and report links. Limits: 50 emails a day per wallet, 10 a day to the same address, at most 3 links (no link shorteners or raw-IP links). Refused sends (not yet accepted, blocked, over a limit) are not charged. $0.005 per send. GET with query parameters or POST JSON {to, subject, body}. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Opt-in email from your agent
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • translateTranslate up to 5,000 characters between English and 49 languages, or between any two of them (via English), with automatic source-language detection.Reads text others writeCan send data out1 finding

    Why it is in this group

    • The input addresses people (input.to): what it sends cannot be taken back.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a destination (input.to): what the tool writes goes there.

    Findings

    • Warn

      The server marks "translate" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.

      What this check looks for
    Description, as the server wrote it
    Translate up to 5,000 characters between English and 49 languages, or between any two of them (via English), with automatic source-language detection. Runs on our own server (open-source Argos models), so your text is not sent to a third-party translation service. Every model passes our quality check, and nonsense output is refused and not charged. Quality is best for major languages and rougher for less common ones. $0.01 per call. Price: $0.01 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Translate text (private, self-hosted)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
Changes things 19Claude Code: ask
  • certsUnexpired TLS certificates publicly logged for a domain (Certificate Transparency, via crt.sh with Cert Spotter as backup): subdomains (up to 500) and wildcards found, the 25 newes…Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Unexpired TLS certificates publicly logged for a domain (Certificate Transparency, via crt.sh with Cert Spotter as backup): subdomains (up to 500) and wildcards found, the 25 newest certificates, issuing authorities with counts, certificates issued in the last 7 days, expiry dates, and flags such as sudden new issuance. Useful for mapping a domain's attack surface or spotting look-alike infrastructure. Price: $0.01 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Certificate Transparency subdomain finder
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • dnsDNS records (A, AAAA, MX, TXT, NS, CNAME, SOA) from public resolvers (Cloudflare, with Google and Quad9 as fallbacks), the domain's IP addresses, the network owner of its first IP …Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    DNS records (A, AAAA, MX, TXT, NS, CNAME, SOA) from public resolvers (Cloudflare, with Google and Quad9 as fallbacks), the domain's IP addresses, the network owner of its first IP (ASN, network name, prefix, country), its HTTP response headers, and a ping summary. One call replaces several lookups. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: DNS and host lookup
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • domainriskThousands of email domains that are risky to send to: throwaway (disposable) providers still accepting mail, catch-all domains, and domains that can't receive mail.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Thousands of email domains that are risky to send to: throwaway (disposable) providers still accepting mail, catch-all domains, and domains that can't receive mail. Each entry has risk level, confidence, the date it was last confirmed and why it is listed. Built daily from DNS checks and real verification results. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Daily risky email-domain feed
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • email_verifyIs this email address real and safe to send to?Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Is this email address real and safe to send to? Returns status (deliverable, undeliverable, risky, unknown, invalid), reason, and 0-100 score. Checks syntax, domain and MX records, asks the mail server whether the mailbox exists (no email is sent), detects catch-all domains, disposable/throwaway providers, role accounts (info@, support@), free providers, and suggests typo fixes (gmial.com -> gmail.com). Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Email verification
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • image_editEdit an uploaded image with a JSON list of ops, always re-rendered from the original (no quality loss), with numbered versions to undo or branch (from=0 = original).Reads text others writeNo findings

    Why it is in this group

    • The name "image_edit" says "edit": a word for a change.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Edit an uploaded image with a JSON list of ops, always re-rendered from the original (no quality loss), with numbered versions to undo or branch (from=0 = original). Resize, crop, social presets (instagram_square, story, og_image, youtube_thumbnail, favicon...), rotate, adjust colors, blur, captions, watermarks, logo overlay, rounded/circle. Output jpeg/png/webp/avif/gif/ico, max_kb size target. Metadata stripped. FREE TIER: add free=1. Price: $0.002 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Non-destructive image editor
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • image_infoFor an uploaded image: original details plus every version (ops used, size, format) with fresh download links.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    For an uploaded image: original details plus every version (ops used, size, format) with fresh download links. FREE TIER: add free=1. Price: $0.002 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Image versions and fresh links
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • image_uploadUpload an image once (url=, or POST the bytes) and get an image_id for /image-edit.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "image_upload" says "upload": a word for a change.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Upload an image once (url=, or POST the bytes) and get an image_id for /image-edit. Every edit is re-rendered from this original, so quality never drifts. Returns size, format, colors, metadata, GPS warning, and a clean copy link. The upload is kept 7 days after you last use it. FREE TIER: add free=1. Price: $0.002 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Image upload for lossless editing
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • markdownAny public URL as clean Markdown: main content only (no nav, ads, footers, scripts), with headings, links, lists, tables, and code kept.Reads text others writeCan send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Any public URL as clean Markdown: main content only (no nav, ads, footers, scripts), with headings, links, lists, tables, and code kept. Handles articles, docs, plain-text files and PDFs (text-based PDFs only, first 50 pages). Returns title, author, date, word count, and approx token count. Optional max_chars (default 20000, max 100000). Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Web page to clean Markdown
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • metaStructured facts a page declares about itself, so you don't have to read the page: highlights (product price, currency, stock, rating, brand; event dates, venue, address; article h…Reads text others writeCan send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Structured facts a page declares about itself, so you don't have to read the page: highlights (product price, currency, stock, rating, brand; event dates, venue, address; article headline, author, date; organization, recipe, job posting, FAQ), link preview (title, description, image), canonical URL, language, plus raw OpenGraph, Twitter card, JSON-LD (broken blocks repaired), and microdata. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Page structured data and preview
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • onchainEthereum and Base right now: latest block, average block time, transactions per block, how full blocks are, next base fee and typical priority fee (gwei).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Ethereum and Base right now: latest block, average block time, transactions per block, how full blocks are, next base fee and typical priority fee (gwei). Refreshed about every 30 minutes; taken offline (not sold) if it ever goes 2 hours without a good update. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: On-chain snapshot: Ethereum and Base (every 30 min)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • publicdataTop 10 cryptocurrencies by market cap: USD price, market cap, rank and 24h change in clean JSON.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Top 10 cryptocurrencies by market cap: USD price, market cap, rank and 24h change in clean JSON. Refreshed about every 30 minutes (CoinGecko, with CoinPaprika as backup); taken offline (not sold) if it goes 2 hours without a good update. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Top 10 cryptocurrencies (every 30 min)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • redirectsFollows a URL hop by hop (up to 15) without loading the final page's content: status, Location, timing and server for each hop, final URL, and flags for loops, too many hops, HTTPS…Reads text others writeCan send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Follows a URL hop by hop (up to 15) without loading the final page's content: status, Location, timing and server for each hop, final URL, and flags for loops, too many hops, HTTPS-to-HTTP downgrades, cross-domain jumps, suspected open redirects, and meta-refresh redirects. Use on short links or tracking links before opening them. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Redirect chain tracer
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • robotsWhether a URL may be fetched by 20 named AI and search agents (GPTBot, ChatGPT-User, ClaudeBot, Claude-User, PerplexityBot, Google-Extended, Googlebot, Bingbot, CCBot, and more), p…Reads text others writeCan send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "crawl": it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Whether a URL may be fetched by 20 named AI and search agents (GPTBot, ChatGPT-User, ClaudeBot, Claude-User, PerplexityBot, Google-Extended, Googlebot, Bingbot, CCBot, and more), parsed to RFC 9309: the rules group that applied, crawl-delay, sitemaps, Content-Signal preferences, and whether llms.txt or ai.txt exist. Check before crawling, or audit a site's AI policy. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: AI crawler access checker (robots.txt)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • rssLatest items from a site's RSS, Atom, or JSON feed.Reads text others writeCan send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Latest items from a site's RSS, Atom, or JSON feed. Give a feed URL or any page: the feed is found automatically. Returns clean titles, links, ISO dates, authors, short plain-text summaries, and categories. Cheaper and more reliable than scraping a blog or news page for new posts. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: RSS / Atom feed reader
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • secA US public company's SEC filings in one call: company facts (industry, state, fiscal year end, exchanges), the latest 10-K, 10-Q, 8-K, proxy (DEF 14A) and S-1 (or 20-F and 6-K for…Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    A US public company's SEC filings in one call: company facts (industry, state, fiscal year end, exchanges), the latest 10-K, 10-Q, 8-K, proxy (DEF 14A) and S-1 (or 20-F and 6-K for foreign companies) with direct document links, how many 8-Ks were filed in the last 90 days, and the most recent filings. By ticker or CIK, straight from SEC EDGAR. Price: $0.01 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: SEC EDGAR company filings
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • site_checkOne-call safety and setup audit for a domain: verdict (score + grade + summary) and a ranked list of issues, plus DNS, email security (SPF, DKIM on common selectors, DMARC, MTA-STS…Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    One-call safety and setup audit for a domain: verdict (score + grade + summary) and a ranked list of issues, plus DNS, email security (SPF, DKIM on common selectors, DMARC, MTA-STS), SSL certificate (trust, issuer, expiry, days left, SANs, TLS version), domain age and registrar (RDAP), and HTTP redirect chain with security headers. Flags scam signals like newly registered domains and untrusted certificates. Price: $0.01 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Domain safety and setup audit
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • testIs this URL or API working right now?Reads text others writeCan send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Is this URL or API working right now? Returns ok/down, HTTP status, time to first byte and total time, redirects followed, size, content type, server, HTTPS and security headers, the page title, and for JSON APIs whether the response is valid JSON and its top-level keys. A site that is down is an answer, not an error. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: URL and API health check
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • tokenPrice snapshot for a crypto token by symbol, or by contract address on Ethereum or Base: USD price, 24h change, market cap, 24h volume, and market-cap rank, from CoinGecko.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Price snapshot for a crypto token by symbol, or by contract address on Ethereum or Base: USD price, 24h change, market cap, 24h volume, and market-cap rank, from CoinGecko. Unknown tokens and a busy price service are refused and not charged. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Crypto token snapshot
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • walletOn-chain overview for any Base address: native balance and nonce.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    On-chain overview for any Base address: native balance and nonce. One call replaces multiple RPC queries. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Base wallet overview
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
Only reads 3Show 3 read-only tools
  • diffWhat changed on a page.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "diff" says "diff": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    What changed on a page. First call returns a snapshot_id; call again with since=<snapshot_id> to get only the added, removed, and changed lines (e.g. a price going from $29 to $39) instead of re-reading the page. Snapshots last 30 days. Or pass compare=<second url> to diff two pages directly. Works on HTML, PDFs, and text. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Page change monitor
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • fetchFetches a public URL and returns its visible text (up to 15,000 characters) with title, meta description, status code, and final URL after redirects.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "fetch" says "fetch": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • untrusted-inputThe description's first sentence says "fetch" and "url": it reads web pages, which anyone can write.
    • untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

    Claude Code: ask, because it is tagged outbound.

    Description, as the server wrote it
    Fetches a public URL and returns its visible text (up to 15,000 characters) with title, meta description, status code, and final URL after redirects. Scripts, styles, navigation, headers, footers, sidebars and forms are removed. For a cleaner main-content extract with headings and links kept, use /markdown. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: Web page plain text
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • github_treeCompact map of a public GitHub repository for reading code with an LLM: filtered file tree (node_modules, build output and lock files removed), entry points, key config files, larg…Reads private dataReads text others writeNo findings

    Why it is in this group

    • The name "github_tree" says "tree": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "repository": it reads repositories.
    • private-dataThe input names a repo (input.repo), and the tool reads: it reads repositories.
    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Compact map of a public GitHub repository for reading code with an LLM: filtered file tree (node_modules, build output and lock files removed), entry points, key config files, largest files with token estimates, total token estimate, languages, stars, license, last push, and the start of the README. Decide what to read before spending tokens. Price: $0.005 USDC on Base per successful call (x402).

    Annotations the server declares

    • title: GitHub repository map for LLMs
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Exfil surface

Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.

297 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 1 private-data tool breaks every one.

Reads private data1
Reads files, mail, messages, databases, secrets, calendars or repositories.
Reads text others write27
Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
Can send data out11
Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
deny to break every path
{
  "permissions": {
    "deny": [
      "mcp__tools__github_tree"
    ]
  }
}

Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.

Remote posture

What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.

Tools listed without sign-in

Looked at Nov 17, 2026, 15:18 UTC, unchanged since Nov 17, 2026. No credentials were sent and no tool was called. What is looked at.

Reachable
yes
HTTP status of an anonymous initialize
200
tools/list served without credentials
yes
Tools that change things, listed without credentials
17
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
not observed
Authorization servers
not observed
Legacy HTTP+SSE transport
no
Dynamic client registration
not observed
Client ID metadata documents
not observed
  • WarnThe whole server

    The endpoint served tools/list without credentials, and the list has 17 tools that change things (17 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://data.killbounce.app/mcp: tools/list returned 20 tools to an anonymous client, including email_verify (write), site_check (write), dns (write), markdown (write), meta (write) and 12 more

    What this check looks for

Findings

Matches of fixed checks, with the text they matched.

1 warn
  • The server marks "translate" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.

    What this check looks for

History

Changes to its tool definitions and to its remote endpoint, newest first.

  1. Tools7 added, 16 changed.+7 added~16 changedView diff
    • Added:codeDeletes or cancels
    • Added:email_sendCan't be undone
    • Added:fileDeletes or cancels
    • Added:kvDeletes or cancels
    • Added:secChanges things
    • Added:testChanges things
    • Added:translateCan't be undone
    • Changed:certsChanges thingsdescription
    • Changed:diffOnly readsinputSchema
    • Changed:dnsChanges thingsdescription
    • Changed:domainriskChanges thingsdescription
    • Changed:fetchOnly readsdescription
    • Changed:github_treeOnly readsinputSchema
    • Changed:image_editChanges thingsinputSchema
    • Changed:image_infoChanges thingsinputSchema
    • Changed:image_uploadChanges thingsdescriptioninputSchema
    • Changed:markdownChanges thingsdescription
    • Changed:metaChanges thingsdescription
    • Changed:onchainChanges thingstitledescriptionannotations
    • Changed:publicdataChanges thingstitledescriptionannotations
    • Changed:rssChanges thingsinputSchema
    • Changed:site_checkChanges thingsdescription
    • Changed:tokenChanges thingsdescriptioninputSchema

Install

Pinned to version 1.2.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http tools https://data.killbounce.app/mcp
claude mcp add --transport http tools https://data.killbounce.app/mcp

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json
{
  "permissions": {
    "allow": [
      "mcp__tools__github_tree"
    ],
    "ask": [
      "mcp__tools__email_verify",
      "mcp__tools__site_check",
      "mcp__tools__dns",
      "mcp__tools__markdown",
      "mcp__tools__meta",
      "mcp__tools__diff",
      "mcp__tools__fetch",
      "mcp__tools__token",
      "mcp__tools__wallet",
      "mcp__tools__certs",
      "mcp__tools__robots",
      "mcp__tools__rss",
      "mcp__tools__redirects",
      "mcp__tools__image_upload",
      "mcp__tools__image_edit",
      "mcp__tools__image_info",
      "mcp__tools__sec",
      "mcp__tools__test",
      "mcp__tools__email_send",
      "mcp__tools__translate",
      "mcp__tools__onchain",
      "mcp__tools__domainrisk",
      "mcp__tools__publicdata"
    ],
    "deny": [
      "mcp__tools__kv",
      "mcp__tools__file",
      "mcp__tools__code"
    ]
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as tools. Review it before you use it.

Capability tags behind 27 rules
  • mcp__tools__email_verify untrusted-input
  • mcp__tools__site_check untrusted-input
  • mcp__tools__dns untrusted-input
  • mcp__tools__markdown untrusted-input, outbound
  • mcp__tools__meta untrusted-input, outbound
  • mcp__tools__diff untrusted-input, outbound
  • mcp__tools__fetch untrusted-input, outbound
  • mcp__tools__token untrusted-input
  • mcp__tools__wallet untrusted-input
  • mcp__tools__certs untrusted-input
  • mcp__tools__github_tree private-data, untrusted-input
  • mcp__tools__robots untrusted-input, outbound
  • mcp__tools__rss untrusted-input, outbound
  • mcp__tools__redirects untrusted-input, outbound
  • mcp__tools__image_upload untrusted-input, outbound
  • mcp__tools__image_edit untrusted-input
  • mcp__tools__image_info untrusted-input
  • mcp__tools__sec untrusted-input
  • mcp__tools__test untrusted-input, outbound
  • mcp__tools__kv untrusted-input
  • mcp__tools__file untrusted-input
  • mcp__tools__code untrusted-input
  • mcp__tools__email_send untrusted-input, outbound
  • mcp__tools__translate untrusted-input, outbound
  • mcp__tools__onchain untrusted-input
  • mcp__tools__domainrisk untrusted-input
  • mcp__tools__publicdata untrusted-input

Pin

Keep the decision: fail CI when the tools change, follow every change, or show a badge.

Pin in CI

Fail a build when the tools you reviewed are not the tools the server serves today.

curl -s "https://radar.anlyon.com/api/v1/servers/app.killbounce.data/tools/verify?hash=d452fb722a8500913262dfc13f09df2e28e14e836a3bf0c72f2e61f9d08e9c22" | jq -e '.changed == false'

Pinned hash d452fb722a85, from Nov 21, 2026, 08:17 UTC.

Watch

Get told when its tools change: a new tool, a new description, a new class.

https://radar.anlyon.com/servers/app.killbounce.data/tools/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Badge

Anlyon Radar badge for app.killbounce.data/tools
[![Anlyon Radar](https://radar.anlyon.com/badge/app.killbounce.data/tools.svg)](https://radar.anlyon.com/servers/app.killbounce.data/tools)

Source: MCP Registry. Synced Nov 21, 2026, 08:17 UTC.

KillBounce Data MCP server: questions and answers

Answered from the scan above, not from the publisher's description.

Is the KillBounce Data MCP server safe?

No scan can promise that. What its definitions show: 27 tools, 24 of which can change, send or delete something, and 1 finding from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.

Can the KillBounce Data MCP server delete data?

Yes. 3 tools can delete, cancel or revoke something: kv, file and code.

Can the KillBounce Data MCP server do things that cannot be undone?

Yes. 2 tools can send, pay, publish or merge: email_send and translate.

What tools does the KillBounce Data MCP server have?

27 tools: 3 destructive, 2 irreversible, 19 write, 3 read. kv, file, code, email_send, translate, email_verify, site_check, dns and 19 more.

Does the KillBounce Data MCP server need sign-in or OAuth?

Its remote endpoint listed its tools to an anonymous client, without sign-in.

Does the KillBounce Data MCP server send data to outside services?

11 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): markdown, meta, diff, fetch, robots, rss and 5 more.

Has the KillBounce Data MCP server changed its tools?

Yes, once since Anlyon Radar started tracking it, most recently on Nov 21, 2026. Every change is on this page and in its Atom feed.