Skip to content

FoxForm MCP server

io.github.vitorrocha-panda/foxformvitorrocha-pandastreamable-httpv1.0.0Scanned

Build, publish and read scored forms and quizzes where the score picks the next screen.

10 tools: 2 can delete or cancel things, 1 do something that can't be undone, 2 change things and 5 only read.

Tools: 2 delete, 1 irreversible, 2 write, 5 read.
  • Reads private data · 0
  • Reads text others write · 10
  • Can send data out · 2
  • 3 findings

Tools listed without sign-in. Remote posture

With Anlyon

2 tools here can delete or cancel things and 1 does 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 FoxForm MCP server can do

The FoxForm MCP server is published by the GitHub account vitorrocha-panda. It has 10 tools. It can delete or cancel things (2 tools: foxform_update_form and foxform_unpublish_form), do things that cannot be undone, such as sending, paying or publishing (foxform_publish_form) and create or change data (2 tools: foxform_export_responses and foxform_create_form). 5 tools only read.

10 tools bring in text other people can write (foxform_list_forms, foxform_get_form, foxform_list_responses and 7 more); 2 tools can send data out of your system (foxform_create_form and foxform_publish_form).

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 once since Anlyon Radar first saw them.

Last scanned Nov 21, 2026, version 1.0.0.

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.

10 tools

Deletes or cancels 2Claude Code: deny
  • foxform_unpublish_formUnpublish a form (takes it offline; stops accepting responses).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
    Unpublish a form (takes it offline; stops accepting responses). Requires a WRITE-scoped API key. Args: - form_id (string) Returns: a confirmation message.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: true
    • openWorldHint: true
  • foxform_update_formUpdate an existing form's fields, including each screen's conditional logic (branching).Reads text others write1 finding

    Why it is in this group

    • The name "foxform_update_form" says "update": 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.

    Findings

    • Info

      The description is 4,725 characters long, over the 2,000 this check looks for.

      Matched text

      description ends: "...`/`conditions`/`goto`/`jump`/`nextScreen` anywhere, are not read by any renderer."

      What this check looks for
    Description, as the server wrote it
    Update an existing form's fields, including each screen's conditional logic (branching). Requires a WRITE-scoped API key. Only the fields you pass are changed. Args: - form_id (string): the form to update (required) - title (string, optional) - description (string, optional) - theme (string, optional): one of amber|base|midnight|ocean|sunset|forest|lavender|minimal - questions (array, optional): replaces the FULL screen list — there is no per-screen patch. To add logic to one screen: call foxform_get_form with response_format 'json', edit that screen's `logic`, and send the whole array back. - thank_you_message (string, optional) - format ('single_question' | 'single_page' | 'chatbot', optional) - sections (array, optional): single_page page groups, [{ id }] - chatbot_skin ('instagram' | 'whatsapp', optional) - chatbot_profile (object, optional): chatbot header profile ({ name, handle, followers_text, phone, status, verified, avatar_url }) Returns: { form } with the updated form. Screen fields are validated: unknown fields are REJECTED instead of being stored and ignored (the API accepts arbitrary keys but no renderer reads them), `then.targetScreenId` must be the id of a screen in the same payload, and `{{variables}}` that no screen exposes come back as warnings. Form format (DEVF-296 / DEVF-322): pass `format` to choose how the form is presented. - 'single_question' (default): one screen at a time (classic). - 'single_page': every screen on one scrollable page. Optionally group screens into pages with `sections`: `[{ id }]` in order (the 'Section N' label comes from position; no editable title). - 'chatbot': an Instagram/WhatsApp DM-style conversation. Set `chatbot_skin` ('instagram' | 'whatsapp', default instagram) and `chatbot_profile` (all fields optional, each with a fallback). The question/branching engine is unchanged — only the presentation differs. These fields require the account's form-format feature to be enabled (live in production); when off the API ignores them and the form stays single_question. End screen ('end'): title + description, an optional button (buttonEnabled false hides it; buttonText, buttonLink, buttonOpenNewTab) and an optional automatic redirect after submit: autoRedirect true + autoRedirectUrl (absolute http/https URL; the form link's query params such as utm_* are passed along and {{param}} tokens are filled from them) + autoRedirectDelay (seconds 0-10, default 3). The response is recorded before the redirect; respondents see a "Redirecting in N s" line with a "Go now" link. CONDITIONAL LOGIC (branching), per screen — stored in `questions[].logic`: logic.conditionalNavigationV2 = { enabled: true, groups: [ // groups are OR-joined; FIRST matching group wins { id: "grp-1", conditions: [ // conditions inside a group are AND-joined { id: "cond-1", left: "{{quer_testar}}", operator: "equal_to", right: "Ainda não" } ], then: { type: "specific_screen", targetScreenId: "s-motivos" } } ] } - `then.type`: 'next_screen' | 'previous_screen' | 'specific_screen' (needs targetScreenId = another screen's `id`) | 'end_form'. Add `then.url` (+ optional `openNewTab`) to redirect to an external URL instead. - `operator`: 'equal_to' | 'not_equal_to' | 'greater_than' | 'greater_or_equal_than' | 'less_than' | 'less_or_equal_than' | 'contains'. - `left`/`right` are EXPRESSION strings: a literal ("10", "Ainda não"), a variable ("{{score}}", "{{minha_var}}" = the screen's `variableName`), or arithmetic ("calc({{peso}}/(({{altura}}/100)*({{altura}}/100)))"). - Comparing an ANSWER: use `left: "{{<variableName of the deciding screen>}}"` and `right` = the option's `label` OR its `value` (both match). - `{{score}}` is the running sum of `points` on the options picked so far (`choices[].points`, `images[].points`) — that is how score-based branching works. - A navigation group with no conditions NEVER matches. `enabled: false` stores the rules but disables them. - Screen-level conditional display uses the same group shape: `logic.display = { enabled: true, groups: [...], showAfterSeconds?: n }` (`then` is ignored — THEN means "show"). - Other logic keys: `logic.autoAdvance = { enabled, delaySeconds? }`, `logic.navigationBehavior = { onButtonClick?, onAutoAdvance?, targetScreenId? }`. - `logic.conditionalNavigation` (legacy, pre-DEVF-161) is still read and migrated on load — don't author new rules with it. Unknown fields are REJECTED (they used to be stored and silently ignored): `logic` as an array, or `rules`/`branching`/`conditions`/`goto`/`jump`/`nextScreen` anywhere, are not read by any renderer.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
Can't be undone 1Claude Code: ask
  • foxform_publish_formPublish a form so it's live at its public URL and can accept responses.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "foxform_publish_form" says "publish": a word for a change that another call does not undo.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe name "foxform_publish_form" says "publish": it sends or publishes something to other people.
    Description, as the server wrote it
    Publish a form so it's live at its public URL and can accept responses. Requires a WRITE-scoped API key. Args: - form_id (string) Returns: { form, public_url } — the published form (status 'published') and its live public link `https://forms.foxform.app/<slug>`. When sharing the link with the user, use `public_url` EXACTLY as returned; do NOT build the URL yourself (the public domain is forms.foxform.app, NOT foxform.app or the API/app domain). (May fail with a plan-limit error on Free accounts.)

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
Changes things 2Claude Code: ask
  • foxform_create_formCreate a new form, including per-screen conditional logic.Reads text others writeCan send data out1 finding

    Why it is in this group

    • The name "foxform_create_form" says "create": 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 destination (input.chatbot_profile.phone): what the tool writes goes there.

    Findings

    • Info

      The description is 4,625 characters long, over the 2,000 this check looks for.

      Matched text

      description ends: "...`/`conditions`/`goto`/`jump`/`nextScreen` anywhere, are not read by any renderer."

      What this check looks for
    Description, as the server wrote it
    Create a new form, including per-screen conditional logic. Requires a WRITE-scoped API key. Args: - title (string): form title (required) - description (string, optional) - theme (string, optional): one of amber|base|midnight|ocean|sunset|forest|lavender|minimal (default amber, the copper-to-amber FoxForm theme) - questions (array, optional): array of screen objects ({ id, type, title, required, variableName?, choices?, logic?, ... }); omit to start empty - thank_you_message (string, optional) - format ('single_question' | 'single_page' | 'chatbot', optional): how the form is presented (see below) - sections (array, optional): single_page page groups, [{ id }] - chatbot_skin ('instagram' | 'whatsapp', optional): chatbot chrome - chatbot_profile (object, optional): chatbot header profile ({ name, handle, followers_text, phone, status, verified, avatar_url }) Returns: { form } with the created form (including its id and slug). The form starts as a draft — call foxform_publish_form to make it live. Screen fields are validated: unknown fields are REJECTED instead of being stored and ignored, and branching rules are cross-checked against the screen ids in the same payload. Form format (DEVF-296 / DEVF-322): pass `format` to choose how the form is presented. - 'single_question' (default): one screen at a time (classic). - 'single_page': every screen on one scrollable page. Optionally group screens into pages with `sections`: `[{ id }]` in order (the 'Section N' label comes from position; no editable title). - 'chatbot': an Instagram/WhatsApp DM-style conversation. Set `chatbot_skin` ('instagram' | 'whatsapp', default instagram) and `chatbot_profile` (all fields optional, each with a fallback). The question/branching engine is unchanged — only the presentation differs. These fields require the account's form-format feature to be enabled (live in production); when off the API ignores them and the form stays single_question. End screen ('end'): title + description, an optional button (buttonEnabled false hides it; buttonText, buttonLink, buttonOpenNewTab) and an optional automatic redirect after submit: autoRedirect true + autoRedirectUrl (absolute http/https URL; the form link's query params such as utm_* are passed along and {{param}} tokens are filled from them) + autoRedirectDelay (seconds 0-10, default 3). The response is recorded before the redirect; respondents see a "Redirecting in N s" line with a "Go now" link. CONDITIONAL LOGIC (branching), per screen — stored in `questions[].logic`: logic.conditionalNavigationV2 = { enabled: true, groups: [ // groups are OR-joined; FIRST matching group wins { id: "grp-1", conditions: [ // conditions inside a group are AND-joined { id: "cond-1", left: "{{quer_testar}}", operator: "equal_to", right: "Ainda não" } ], then: { type: "specific_screen", targetScreenId: "s-motivos" } } ] } - `then.type`: 'next_screen' | 'previous_screen' | 'specific_screen' (needs targetScreenId = another screen's `id`) | 'end_form'. Add `then.url` (+ optional `openNewTab`) to redirect to an external URL instead. - `operator`: 'equal_to' | 'not_equal_to' | 'greater_than' | 'greater_or_equal_than' | 'less_than' | 'less_or_equal_than' | 'contains'. - `left`/`right` are EXPRESSION strings: a literal ("10", "Ainda não"), a variable ("{{score}}", "{{minha_var}}" = the screen's `variableName`), or arithmetic ("calc({{peso}}/(({{altura}}/100)*({{altura}}/100)))"). - Comparing an ANSWER: use `left: "{{<variableName of the deciding screen>}}"` and `right` = the option's `label` OR its `value` (both match). - `{{score}}` is the running sum of `points` on the options picked so far (`choices[].points`, `images[].points`) — that is how score-based branching works. - A navigation group with no conditions NEVER matches. `enabled: false` stores the rules but disables them. - Screen-level conditional display uses the same group shape: `logic.display = { enabled: true, groups: [...], showAfterSeconds?: n }` (`then` is ignored — THEN means "show"). - Other logic keys: `logic.autoAdvance = { enabled, delaySeconds? }`, `logic.navigationBehavior = { onButtonClick?, onAutoAdvance?, targetScreenId? }`. - `logic.conditionalNavigation` (legacy, pre-DEVF-161) is still read and migrated on load — don't author new rules with it. Unknown fields are REJECTED (they used to be stored and silently ignored): `logic` as an array, or `rules`/`branching`/`conditions`/`goto`/`jump`/`nextScreen` anywhere, are not read by any renderer.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • foxform_export_responsesExport all responses for a form as CSV text (one row per response, columns = questions).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
    Export all responses for a form as CSV text (one row per response, columns = questions). Large exports are truncated — use foxform_list_responses with pagination for very large datasets. Args: - form_id (string) Returns: raw CSV text.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
Only reads 5Show 5 read-only tools
  • foxform_get_formFetch a single form by ID, including its full question list, per-screen conditional logic and settings.Reads text others write1 finding

    Why it is in this group

    • The name "foxform_get_form" says "get": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.

    Findings

    • Info

      The description is 4,411 characters long, over the 2,000 this check looks for.

      Matched text

      description ends: "...`/`conditions`/`goto`/`jump`/`nextScreen` anywhere, are not read by any renderer."

      What this check looks for
    Description, as the server wrote it
    Fetch a single form by ID, including its full question list, per-screen conditional logic and settings. Args: - form_id (string): the form's ID (from foxform_list_forms) - response_format ('markdown' | 'json') Returns { form, public_url }: the full form object (id, title, description, slug, status, theme, questions[], thank_you_message, timestamps, plus `format` and — when set — `sections`, `chatbot_skin`, `chatbot_profile`) plus `public_url` — the public link (`https://forms.foxform.app/<slug>`), which is live only when status is 'published'. Use `public_url` verbatim; never build the URL yourself (the public domain is forms.foxform.app, not foxform.app). Each screen in `questions[]` carries its own `logic` (branching / conditional display) and, for choice screens, `choices[]`/`images[]` with their `points` and `value`. The markdown output summarises every rule; use response_format 'json' to get the exact stored objects (that's the shape foxform_update_form expects back). Form format (DEVF-296 / DEVF-322): pass `format` to choose how the form is presented. - 'single_question' (default): one screen at a time (classic). - 'single_page': every screen on one scrollable page. Optionally group screens into pages with `sections`: `[{ id }]` in order (the 'Section N' label comes from position; no editable title). - 'chatbot': an Instagram/WhatsApp DM-style conversation. Set `chatbot_skin` ('instagram' | 'whatsapp', default instagram) and `chatbot_profile` (all fields optional, each with a fallback). The question/branching engine is unchanged — only the presentation differs. These fields require the account's form-format feature to be enabled (live in production); when off the API ignores them and the form stays single_question. End screen ('end'): title + description, an optional button (buttonEnabled false hides it; buttonText, buttonLink, buttonOpenNewTab) and an optional automatic redirect after submit: autoRedirect true + autoRedirectUrl (absolute http/https URL; the form link's query params such as utm_* are passed along and {{param}} tokens are filled from them) + autoRedirectDelay (seconds 0-10, default 3). The response is recorded before the redirect; respondents see a "Redirecting in N s" line with a "Go now" link. CONDITIONAL LOGIC (branching), per screen — stored in `questions[].logic`: logic.conditionalNavigationV2 = { enabled: true, groups: [ // groups are OR-joined; FIRST matching group wins { id: "grp-1", conditions: [ // conditions inside a group are AND-joined { id: "cond-1", left: "{{quer_testar}}", operator: "equal_to", right: "Ainda não" } ], then: { type: "specific_screen", targetScreenId: "s-motivos" } } ] } - `then.type`: 'next_screen' | 'previous_screen' | 'specific_screen' (needs targetScreenId = another screen's `id`) | 'end_form'. Add `then.url` (+ optional `openNewTab`) to redirect to an external URL instead. - `operator`: 'equal_to' | 'not_equal_to' | 'greater_than' | 'greater_or_equal_than' | 'less_than' | 'less_or_equal_than' | 'contains'. - `left`/`right` are EXPRESSION strings: a literal ("10", "Ainda não"), a variable ("{{score}}", "{{minha_var}}" = the screen's `variableName`), or arithmetic ("calc({{peso}}/(({{altura}}/100)*({{altura}}/100)))"). - Comparing an ANSWER: use `left: "{{<variableName of the deciding screen>}}"` and `right` = the option's `label` OR its `value` (both match). - `{{score}}` is the running sum of `points` on the options picked so far (`choices[].points`, `images[].points`) — that is how score-based branching works. - A navigation group with no conditions NEVER matches. `enabled: false` stores the rules but disables them. - Screen-level conditional display uses the same group shape: `logic.display = { enabled: true, groups: [...], showAfterSeconds?: n }` (`then` is ignored — THEN means "show"). - Other logic keys: `logic.autoAdvance = { enabled, delaySeconds? }`, `logic.navigationBehavior = { onButtonClick?, onAutoAdvance?, targetScreenId? }`. - `logic.conditionalNavigation` (legacy, pre-DEVF-161) is still read and migrated on load — don't author new rules with it. Unknown fields are REJECTED (they used to be stored and silently ignored): `logic` as an array, or `rules`/`branching`/`conditions`/`goto`/`jump`/`nextScreen` anywhere, are not read by any renderer.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • foxform_get_form_analyticsAggregated analytics for a form: overview KPIs (total responses, form views, response rate, avg response time), the responses-over-time timeline, and per-question stats.Reads text others writeNo findings

    Why it is in this group

    • The name "foxform_get_form_analytics" says "get": a word for reading.

    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
    Aggregated analytics for a form: overview KPIs (total responses, form views, response rate, avg response time), the responses-over-time timeline, and per-question stats. Args: - form_id (string) - response_format ('markdown' | 'json') Returns: { overview, timeline: [{date,count}], questions: [...], response_time_distribution?, views_timeline? }. Note: form_views / response_rate are null until the form has tracked views (forward-only).

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • foxform_get_responseFetch one response by its ID (from foxform_list_responses).Reads text others writeNo findings

    Why it is in this group

    • The name "foxform_get_response" says "get": a word for reading.

    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
    Fetch one response by its ID (from foxform_list_responses). Args: - response_id (string) - response_format ('markdown' | 'json') Returns the full response object (id, form_id, answers, submitted_at, metadata).

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • foxform_list_formsList the forms owned by the authenticated FoxForm account, newest first.Reads text others writeNo findings

    Why it is in this group

    • The name "foxform_list_forms" says "list": a word for reading.

    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
    List the forms owned by the authenticated FoxForm account, newest first. Args: - page (number): 1-based page number (default 1) - limit (number): page size, 1-100 (default 20) - response_format ('markdown' | 'json'): output format (default markdown) Returns: { total, page, limit, count, forms: [{ id, title, status, slug, public_url, questions_count, updated_at }] } The public link of a form is `public_url` (`https://forms.foxform.app/<slug>`) — always use this exact value, do NOT build the URL yourself (the public domain is forms.foxform.app, not foxform.app). `public_url` is only live when `status` is 'published'. Use this first to discover form IDs, then call foxform_get_form / foxform_get_form_analytics.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • foxform_list_responsesList submitted responses for a form, newest first.Reads text others writeNo findings

    Why it is in this group

    • The name "foxform_list_responses" says "list": a word for reading.

    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
    List submitted responses for a form, newest first. Args: - form_id (string): the form's ID - page (number): 1-based page (default 1) - limit (number): page size, 1-100 (default 20) - response_format ('markdown' | 'json') Returns: { total, page, limit, count, responses: [{ id, submitted_at, answers }] }. For aggregate metrics use foxform_get_form_analytics; for a full dump use foxform_export_responses.

    Annotations the server declares

    • 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.

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

Reads private data0
Reads files, mail, messages, databases, secrets, calendars or repositories.
Can send data out2
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, 19:12 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
5
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
https://mcp.foxform.app/.well-known/oauth-protected-resource
Authorization servers
https://mcp.foxform.app
Legacy HTTP+SSE transport
no
Dynamic client registration
yes
Client ID metadata documents
no
  • HighThe whole server

    The endpoint served tools/list without credentials, and the list has 5 tools that change things (2 destructive, 1 irreversible, 2 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://mcp.foxform.app/mcp: tools/list returned 10 tools to an anonymous client, including foxform_update_form (destructive), foxform_unpublish_form (destructive), foxform_publish_form (irreversible), foxform_export_responses (write), foxform_create_form (write)

    What this check looks for

Findings

Matches of fixed checks, with the text they matched.

3 info
  • The description is 4,411 characters long, over the 2,000 this check looks for.

    Matched text

    description ends: "...`/`conditions`/`goto`/`jump`/`nextScreen` anywhere, are not read by any renderer."

    What this check looks for
  • The description is 4,625 characters long, over the 2,000 this check looks for.

    Matched text

    description ends: "...`/`conditions`/`goto`/`jump`/`nextScreen` anywhere, are not read by any renderer."

    What this check looks for
  • The description is 4,725 characters long, over the 2,000 this check looks for.

    Matched text

    description ends: "...`/`conditions`/`goto`/`jump`/`nextScreen` anywhere, are not read by any renderer."

    What this check looks for

History

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

  1. Tools3 changed.~3 changedView diff

Install

Pinned to version 1.0.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http foxform https://mcp.foxform.app/mcp --header 'x-api-key: <your-x-api-key>'
claude mcp add --transport http foxform https://mcp.foxform.app/mcp --header 'x-api-key: <your-x-api-key>'

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__foxform__foxform_list_forms",
      "mcp__foxform__foxform_get_form",
      "mcp__foxform__foxform_list_responses",
      "mcp__foxform__foxform_get_response",
      "mcp__foxform__foxform_get_form_analytics"
    ],
    "ask": [
      "mcp__foxform__foxform_export_responses",
      "mcp__foxform__foxform_create_form",
      "mcp__foxform__foxform_publish_form"
    ],
    "deny": [
      "mcp__foxform__foxform_update_form",
      "mcp__foxform__foxform_unpublish_form"
    ]
  }
}

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 foxform. Review it before you use it.

Capability tags behind 10 rules
  • mcp__foxform__foxform_list_forms untrusted-input
  • mcp__foxform__foxform_get_form untrusted-input
  • mcp__foxform__foxform_list_responses untrusted-input
  • mcp__foxform__foxform_get_response untrusted-input
  • mcp__foxform__foxform_get_form_analytics untrusted-input
  • mcp__foxform__foxform_export_responses untrusted-input
  • mcp__foxform__foxform_create_form untrusted-input, outbound
  • mcp__foxform__foxform_update_form untrusted-input
  • mcp__foxform__foxform_publish_form untrusted-input, outbound
  • mcp__foxform__foxform_unpublish_form 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/io.github.vitorrocha-panda/foxform/verify?hash=223bbfa5cac07e69050cc0f7eadb36ce10baf0e11aad7d62dda909102bba89af" | jq -e '.changed == false'

Pinned hash 223bbfa5cac0, from Nov 21, 2026, 05:18 UTC.

Watch

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

https://radar.anlyon.com/servers/io.github.vitorrocha-panda/foxform/changes.xml

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

Badge

Anlyon Radar badge for io.github.vitorrocha-panda/foxform
[![Anlyon Radar](https://radar.anlyon.com/badge/io.github.vitorrocha-panda/foxform.svg)](https://radar.anlyon.com/servers/io.github.vitorrocha-panda/foxform)

Source: MCP Registry. Synced Oct 21, 2026, 18:51 UTC.

FoxForm MCP server: questions and answers

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

Is the FoxForm MCP server safe?

No scan can promise that. What its definitions show: 10 tools, 5 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 FoxForm MCP server delete data?

Yes. 2 tools can delete, cancel or revoke something: foxform_update_form and foxform_unpublish_form.

Can the FoxForm MCP server do things that cannot be undone?

Yes. 1 tool can send, pay, publish or merge: foxform_publish_form.

What tools does the FoxForm MCP server have?

10 tools: 2 destructive, 1 irreversible, 2 write, 5 read. foxform_update_form, foxform_unpublish_form, foxform_publish_form, foxform_export_responses, foxform_create_form, foxform_list_forms, foxform_get_form, foxform_list_responses and 2 more.

Does the FoxForm MCP server need sign-in or OAuth?

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

Does the FoxForm MCP server send data to outside services?

2 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): foxform_create_form and foxform_publish_form.

Has the FoxForm MCP server changed its tools?

Yes, once since Anlyon Radar started tracking it, most recently on Nov 7, 2026. Every change is on this page and in its Atom feed.