Skip to content

CurrencyGuard Guard Pricing MCP server

com.currencyguard/guard-pricingcurrencyguard.comstreamable-httpv1.1.1Scanned

FX protection for SMEs — price Guards, live spot/forward rates, settlement dates, and more.

11 tools: 3 change things and 8 only read.

Tools: 0 delete, 0 irreversible, 3 write, 8 read.
  • Reads private data · 0
  • Reads text others write · 0
  • Can send data out · 0
  • 3 findings

Tools listed without sign-in. Remote posture

With Anlyon

3 tools here change things. 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 CurrencyGuard Guard Pricing MCP server can do

The CurrencyGuard Guard Pricing MCP server is published by currencyguard.com, a domain the MCP Registry verified. It has 11 tools. It can create or change data (3 tools: price_guard, net_exposure and request_provider_availability), but cannot delete data. 8 tools only read.

None of its tools reads private data, brings in text other people write, or sends data out.

The fixed checks matched 3 findings (long-description). Each one comes with the text it matched.

Its remote endpoint lists its tools without sign-in, and its tools changed 2 times since Anlyon Radar first saw them.

Last scanned Nov 14, 2026, version 1.1.1.

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.

11 tools

Changes things 3Claude Code: ask
  • net_exposureNet an SME's open foreign-currency bills and invoices, so that a receipt due on or before a payment in the same currency is shown as funding it.1 finding

    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.

    Findings

    • Info

      The description is 2,499 characters long, over the 2,000 this check looks for.

      Matched text

      description ends: "...g a year: confirm with the user and call again. Spec: docs/specs/net_exposure.md."

      What this check looks for
    Description, as the server wrote it
    Net an SME's open foreign-currency bills and invoices, so that a receipt due on or before a payment in the same currency is shown as funding it. Returns the residual on every bill and invoice, a trace of which invoice funds which bill and for how much, per-bucket totals, dependency statements, and series for a ladder chart. It prices nothing: call price_guard on each residual separately, with the item's own due date as settlementDate. Never offsets across currencies. Items in base_currency are ignored. Cancelled items are excluded. Overdue items are listed separately and never netted. Set carry to true only when the client keeps receipts in a wallet or account in that currency, so a surplus can fund a later payment. If you do not know, ask the user whether they hold an account in each foreign currency before calling; do not guess. If their accounting platform is connected, offer to look up its bank accounts and their currencies there, and use what you find. If the answer is "not sure", call the tool twice, carry true and carry false, and show both charts; each carries its own scenario title, and chart_series.scenario holds the same words. THE RESULT HAS TWO PARTS: the JSON, then a PNG of the ladder chart rendered by the service. Show that image to the user as it is. PRESENTING THE RESULT: chart_series.currencies[ccy].open[] is the exposure. Present each open item on its own due date, with its label. Never add open items into one net figure, and never convert them to the base currency at spot. Do not call price_guard or show a Guard price unless the user asks for a price. If the user asks what they can do, lay out three things as facts and stop: do nothing (the amount stays exposed until paid), a forward from a bank or a payment provider (fixes the rate, provider margin inside the rate, an obligation that may need initial margin and margin calls, settled even if the payment does not happen), a Guard (fee known up front, market rate used if better, nothing further to pay if the payment does not happen). Do not rank them or say which fits. Never use the word "hedge". Do not draw a chart of your own from these numbers. DATES: the result carries today, the date the tool used. A due date the user gave without a year means the next such date on or after today, never a past year. If the result carries a warnings[] entry saying every item is before today, the dates are probably missing a year: confirm with the user and call again. Spec: docs/specs/net_exposure.md.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • openWorldHint: false
  • price_guardPrice a Guard — CurrencyGuard's FX protection product for real future payments or receipts.1 finding

    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.

    Findings

    • Info

      The description is 3,183 characters long, over the 2,000 this check looks for.

      Matched text

      description ends: "...ol only when the user asks for a price, a quote, a cost, a fee, or to fix a rate."

      What this check looks for
    Description, as the server wrote it
    Price a Guard — CurrencyGuard's FX protection product for real future payments or receipts. If a tenor like '3 months' is given, call resolve_settlement_date first to get the exact date, then call this tool. THE AMOUNT. Give exactly ONE of foreignAmount and netGuardAmount, never both. Give foreignAmount when the customer has an invoice or bill in the foreign currency and wants to know what it is worth in their own currency: "I have to pay USD 50,000 in March" is foreignAmount 50000, PAY. Give netGuardAmount when the customer names the amount they want in their OWN currency after the fee, and asks what to invoice or how much they can cover: "I need GBP 100,000 from Marco, he pays in euros, what do I quote him?" is netGuardAmount 100000, RECEIVE, guardCurrency GBP, foreignCurrency EUR. With netGuardAmount the server solves for the foreign amount, so the foreignAmount in the response IS THE ANSWER: the amount to invoice (RECEIVE) or the amount covered (PAY). Read it out as the answer to the question. Returns: guardType (Guard-Pay or Guard-Receive), guardRate, guardFee (in the home currency), guardFeePct (a FRACTION: 0.0096 means 0.96%), guardFeePercent (the same figure written as a percentage, e.g. "0.96%" — quote this one), effectiveRate, guardAmount (the foreign amount converted at the Guard rate, in the home currency, WITHOUT the fee), foreignAmount, spotRate, settlementDate, settlementWindowFrom (the first day the Guard rate can be used; the window runs from there to settlementDate. It opens 14 business days before settlementDate on the trading calendar, which is about three weeks in calendar days. Say "14 business days", never "14 days", and read both dates out as given), payLaterCreditFee (the extra for paying the fee at settlement instead of up front. This tool asks for a pay-now quote, so the server returns 0 here. 0 does NOT mean Pay Later is free: say that a small credit fee applies and that it is not in this quote), valid, errors, stale. Also returns netGuardAmount, what the customer ends up with in the home currency once the fee is counted (guardAmount minus guardFee for RECEIVE, guardAmount plus guardFee for PAY), and netGuardAmountNote, that figure in a sentence. Say the sentence. It tells the customer that the fee is paid now, what they get or pay at settlement, and what that nets to. Pilot limit: the Guard amount cannot exceed 1,000,000 in the home currency. Say so before quoting if the amount is clearly above it. The Guard fee is separate money. A payer funds guardAmount plus guardFee; a receiver nets guardAmount minus guardFee. effectiveRate is that all-in figure expressed as a rate. All quotes are indicative: a Guard is bought through the customer's payment provider, and the Guard rate is the rate the provider gives them. Say that when you say "indicative". WHEN NOT TO CALL THIS TOOL. Not for "what can I do", "what are my choices", "is it worth it", or a list of open items. Those get the three things a client can do, with no price for any of them. An amount and a date in such a question do not make it a quote request. Call this tool only when the user asks for a price, a quote, a cost, a fee, or to fix a rate.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • openWorldHint: false
  • request_provider_availabilityFind out whether Guards can be bought through a particular payment provider or bank.No findings

    Why it is in this group

    • The name "request_provider_availability" says "request": a word for a change.
    Description, as the server wrote it
    Find out whether Guards can be bought through a particular payment provider or bank. A Guard is always bought through a payment provider, so this cannot be answered without knowing which one the user pays through — CurrencyGuard checks and replies. ASK BEFORE CALLING. Tell the user plainly that their provider, and their email address if they give one, will be sent to CurrencyGuard so they can answer. Call this only after they say yes. The email address may be left out: the provider on its own is a complete enquiry, and without an address there is simply no one to reply to. Never take an email address from earlier in the conversation — only one the user has just given for this purpose. Never call this more than once in a conversation, and never after the user has declined.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • openWorldHint: false
Only reads 8Show 8 read-only tools
  • compare_guard_with_forwardUse when a user asks whether a Guard is worth it, how it compares with a forward contract from their bank or provider, or what happens if they do nothing.1 finding

    Why it is in this group

    • The name "compare_guard_with_forward" says "compare": a word for reading.

    Findings

    • Info

      The description is 3,365 characters long, over the 2,000 this check looks for.

      Matched text

      description ends: "...ard rate is the rate the provider gives them. Say that when you say "indicative"."

      What this check looks for
    Description, as the server wrote it
    Use when a user asks whether a Guard is worth it, how it compares with a forward contract from their bank or provider, or what happens if they do nothing. ASK BEFORE YOU CALL. The forward side of this comparison should be the customer's own numbers, not CurrencyGuard's guesses. Ask three things first: (1) the forward rate their provider last quoted them for this payment, (2) whether the provider gives them an unsecured line with no deposit, or asks for an initial margin and at what percentage, and (3) at what adverse move the provider demands more cash. Pass what they give you as providerForwardRate, initialMarginPct (0 for an unsecured line) and marginCallThresholdPct. Only call with none of them when the customer cannot answer, and then say in your answer that the forward figures are illustrative defaults. DO NOT RANK. Present every row in the order the output gives it: rates, home-currency amount, cash up front, cash demanded under adverse moves, outcomes under favourable and adverse moves, cancellation, market improvement, worst case. Do not say which one wins, which looks bad, which is cheaper overall, or which the customer should choose. State the numbers. The decision is the customer's. cashNature says what kind of money each figure is: the Guard fee is spent, a forward's deposit and margin calls are held and returned at settlement. Say so when you show them. Held cash is not a cost. outcomes runs each move in both directions. In the favourable rows the Guard lapses and the fee is spent for nothing, and the forward is locked out of the better rate. Show both directions, not only the adverse one. READ OUT comparability.protection AND comparability.pricing beside the cost row, every time. A Guard and a forward are different kinds of protection, and the Guard rate here is mid-market while, bought through a payment provider, it is the rate the provider gives. The cost row is therefore not a like-for-like price comparison. If a tenor like '3 months' is given, call resolve_settlement_date first. DO NOT ALSO CALL price_guard. This returns the Guard rate, fee and amount itself, priced in the same moment as the forward it is compared against. Calling both prices the Guard twice seconds apart, and the market moves between them — the answer then shows two different fees for one Guard. In the cost row, QUOTE THE ALL-IN COST, NOT THE RATE ROW. homeCurrencyAmount.forward already contains the provider margin, because the margin is in the rate, but homeCurrencyAmount.guard does NOT contain the Guard fee. Compare homeCurrencyAmount.guardAllIn with homeCurrencyAmount.forward: those two both carry every cost of their own product. Comparing guard with forward overstates the Guard, by the whole fee. rates.effectiveRate is the same all-in figure expressed as a rate, for setting beside a provider's quoted rate. assumptions.source says, figure by figure, whether the forward side is the customer's data (CUSTOMER, PROVIDER_QUOTE) or ours (DEFAULT, MARKET_CURVE_PLUS_MARGIN). Read assumptions.note out with the answer. Say "your provider's rate" for the customer's figures and "an assumed 5% deposit" for ours. Never present a default as a quote. The Guard figures are indicative: a Guard is bought through the customer's payment provider, and the Guard rate is the rate the provider gives them. Say that when you say "indicative".

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • openWorldHint: false
  • explain_guard_productExplain the Guard product using CurrencyGuard's approved product and FAQ content.No findings

    Why it is in this group

    • The name "explain_guard_product" says "explain": a word for reading.
    Description, as the server wrote it
    Explain the Guard product using CurrencyGuard's approved product and FAQ content. Covers: what the Guard is, how it works, who it is for, how it compares to a forward or to doing nothing, and legal, regulatory, accounting, or eligibility questions. Compare the Guard only with a forward and with doing nothing, as the product text does. If asked to compare it with any other FX product, give the FAQ's definition of what the Guard is and is not, and do not build the comparison: no table, no matching of the Guard fee or the Guard rate to that product's terms.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • openWorldHint: false
  • get_forward_rateUse when you need a forward FX rate.No findings

    Why it is in this group

    • The name "get_forward_rate" says "get": a word for reading.
    Description, as the server wrote it
    Use when you need a forward FX rate. tenor accepts either a period such as 1W, 1M, 3M, 6M, 1Y, or an exact date in ISO form YYYY-MM-DD. Use an exact date whenever you need the forward for the same day a Guard settles: a period resolves to the forward curve's own date, which can be several days away, and a comparison across two different dates is misleading. Rates are taken from the market forward curve, rounded, and for illustration purposes only — not for execution. Returns: forwardRate (indicative mid outright rate for that date), spotMid (indicative spot mid rate for comparison), settlementDate (the date the rate is for). A forwardRate below spotMid means the forward is cheaper than spot for the base currency; above means dearer. Example: base=GBP, quote=USD, tenor=2026-12-29 returns the GBPUSD forward for that day.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • openWorldHint: false
  • get_historic_best_worstAnalyze historic best/worst FX performance for a currency pair over a Guard's duration.No findings

    Why it is in this group

    • The name "get_historic_best_worst" says "get": a word for reading.
    Description, as the server wrote it
    Analyze historic best/worst FX performance for a currency pair over a Guard's duration. Uses Bank of England historic rates. Returns a 'performances' array with 6 entries — BEST and WORST for each of three lookback periods (5, 10, 25 years). Each entry contains: performanceType (BEST or WORST), periodInYears (5, 10 or 25), windowSizeInDays, startDate/endDate, startRate/endRate, performancePercentage, startAmount, finalAmount, deltaAmount (in the window's own money), finalAmountToday, deltaAmountToday (the same move applied to today's amount). For Guard-Pay: negative delta = cost decreased = BEST; positive delta = cost increased = WORST. For Guard-Receive: positive delta = receipt increased = BEST; negative delta = receipt decreased = WORST. When there is a preceding price_guard call, the parameters must match it, including payReceive (PAY or RECEIVE). WHEN TO CALL. Only when the user asks for historic performance, best or worst cases, or historic context. Never to answer "what can I do" or "what would you do". If there is no price_guard quote, leave guardAmount empty; do not call price_guard only to fill it. IF IT FAILS. A result with valid: false has no data. Say that historic figures are unavailable and stop. Do not show the table and do not estimate any figure.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • openWorldHint: false
  • get_spot_rateGet an indicative spot FX rate for a currency pair.No findings

    Why it is in this group

    • The name "get_spot_rate" says "get": a word for reading.
    Description, as the server wrote it
    Get an indicative spot FX rate for a currency pair. Returns a rounded indicative mid rate, for illustration purposes only — not for execution. Example: base=GBP, quote=USD returns the GBPUSD rate.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • openWorldHint: false
  • list_supported_currenciesList the currency pairs currently available for Guard pricing.No findings

    Why it is in this group

    • The name "list_supported_currencies" says "list": a word for reading.
    Description, as the server wrote it
    List the currency pairs currently available for Guard pricing. Always call this rather than assuming which pairs are supported — the list is configuration-driven and changes.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • openWorldHint: false
  • price_guard_extensionGet an indicative price for extending a Guard to a later settlement date.No findings

    Why it is in this group

    • The description's first verb is "get": a word for reading.
    Description, as the server wrote it
    Get an indicative price for extending a Guard to a later settlement date. Use it when a customer asks what it costs to extend a Guard. Returns: extensionCost, what the customer pays to extend, in the home currency. Show it as the answer: "Extending to [date] costs [extensionCost]". It is one figure; there is no breakdown. Also the currencies, the direction, the foreign amount, both dates, valid, errors, stale. Both dates are moved to the next business day when needed, and a note says so. Parameters: guardCurrency, foreignCurrency, payReceive, foreignAmount, originalSettlementDate, extendedSettlementDate. There is no Guard rate input: the server prices the extension at today's market rate, not at the rate of the Guard the customer holds. Say that the figure is indicative for that reason too.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • openWorldHint: false
  • resolve_settlement_dateResolve a relative time period (tenor) to a valid business settlement date for a currency pair.No findings

    Why it is in this group

    • The name "resolve_settlement_date" says "resolve": a word for reading.
    Description, as the server wrote it
    Resolve a relative time period (tenor) to a valid business settlement date for a currency pair. Use this when a user says '3 months', '6 months', '1 year', etc. instead of an exact date. The tenor format is: 1D (days), 1W (weeks), 1M (months), 1Y (years). Examples: '3M' = 3 months, '6M' = 6 months, '1Y' = 1 year. The returned date accounts for weekends and public holidays in both currencies' financial centres. Use the returned settlementDate as the exact date parameter for price_guard.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • openWorldHint: false

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.

No private-data, untrusted-input, outbound path in this server alone: no tool is tagged private-data or untrusted-input or outbound.

Reads private data0
Reads files, mail, messages, databases, secrets, calendars or repositories.
Reads text others write0
Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
Can send data out0
Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.

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 20, 2026, 18:36 UTC, unchanged since Oct 22, 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
3
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 3 tools that change things (3 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://api.currencyguard.com/mcp: tools/list returned 11 tools to an anonymous client, including price_guard (write), net_exposure (write), request_provider_availability (write)

    What this check looks for

Findings

Matches of fixed checks, with the text they matched.

3 info
  • The description is 3,183 characters long, over the 2,000 this check looks for.

    Matched text

    description ends: "...ol only when the user asks for a price, a quote, a cost, a fee, or to fix a rate."

    What this check looks for
  • The description is 3,365 characters long, over the 2,000 this check looks for.

    Matched text

    description ends: "...ard rate is the rate the provider gives them. Say that when you say "indicative"."

    What this check looks for
  • The description is 2,499 characters long, over the 2,000 this check looks for.

    Matched text

    description ends: "...g a year: confirm with the user and call again. Spec: docs/specs/net_exposure.md."

    What this check looks for

History

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

  1. Tools11 changed.~11 changedView diff
  2. Tools1 changed.~1 changedView diff

Install

Pinned to version 1.1.1. Values in angle brackets are yours to fill in.

claude mcp add --transport http guard-pricing https://api.currencyguard.com/mcp
claude mcp add --transport http guard-pricing https://api.currencyguard.com/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__guard-pricing__get_spot_rate",
      "mcp__guard-pricing__list_supported_currencies",
      "mcp__guard-pricing__get_forward_rate",
      "mcp__guard-pricing__compare_guard_with_forward",
      "mcp__guard-pricing__resolve_settlement_date",
      "mcp__guard-pricing__get_historic_best_worst",
      "mcp__guard-pricing__price_guard_extension",
      "mcp__guard-pricing__explain_guard_product"
    ],
    "ask": [
      "mcp__guard-pricing__price_guard",
      "mcp__guard-pricing__net_exposure",
      "mcp__guard-pricing__request_provider_availability"
    ],
    "deny": []
  }
}

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 guard-pricing. Review it before you use it.

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/com.currencyguard/guard-pricing/verify?hash=4e292ec9990c806f5661d49a4fcdd5ca66f8fc9d6f5a3528d2ee9566b2b6a710" | jq -e '.changed == false'

Pinned hash 4e292ec9990c, from Nov 14, 2026, 08:18 UTC.

Watch

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

https://radar.anlyon.com/servers/com.currencyguard/guard-pricing/changes.xml

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

Badge

Anlyon Radar badge for com.currencyguard/guard-pricing
[![Anlyon Radar](https://radar.anlyon.com/badge/com.currencyguard/guard-pricing.svg)](https://radar.anlyon.com/servers/com.currencyguard/guard-pricing)

Source: MCP Registry. Synced Nov 14, 2026, 08:17 UTC.

CurrencyGuard Guard Pricing MCP server: questions and answers

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

Is the CurrencyGuard Guard Pricing MCP server safe?

No scan can promise that. What its definitions show: 11 tools, 3 of which can change, send or delete something, and 3 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.

Can the CurrencyGuard Guard Pricing MCP server delete data?

No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.

Can the CurrencyGuard Guard Pricing MCP server do things that cannot be undone?

No tool is classed as irreversible: none of them sends, pays, publishes or merges, going by their definitions.

What tools does the CurrencyGuard Guard Pricing MCP server have?

11 tools: 3 write, 8 read. price_guard, net_exposure, request_provider_availability, get_spot_rate, list_supported_currencies, get_forward_rate, compare_guard_with_forward, resolve_settlement_date and 3 more.

Does the CurrencyGuard Guard Pricing MCP server need sign-in or OAuth?

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

Does the CurrencyGuard Guard Pricing MCP server send data to outside services?

None of its tools sends, posts or publishes, or takes a recipient, channel or URL.

Has the CurrencyGuard Guard Pricing MCP server changed its tools?

Yes, 2 times since Anlyon Radar started tracking it, most recently on Nov 10, 2026. Every change is on this page and in its Atom feed.