Skip to content

Le Confidant MCP server

ai.leconfidant/le-confidantleconfidant.aistreamable-httpv0.4.0Scanned

Statistics from your own WhatsApp chat analysis as numbers only, plus free WhatsApp text tools.

8 tools: 1 can delete or cancel things, 3 change things and 4 only read.

Tools: 1 delete, 0 irreversible, 3 write, 4 read.
  • Reads private data · 2
  • Reads text others write · 2
  • Can send data out · 1
  • 3 findings

Tools listed without sign-in. Remote posture

With Anlyon

1 tool here can delete or cancel things. Govern it.

  • 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 Le Confidant MCP server can do

The Le Confidant MCP server is published by leconfidant.ai, a domain the MCP Registry verified. It has 8 tools. It can delete or cancel things (1 tool: disconnect_conversation) and create or change data (3 tools: generate_whatsapp_link, format_whatsapp_text and style_whatsapp_text). 4 tools only read.

2 tools read private data (demo_whatsapp_wrapped and get_conversation_snapshot); 2 tools bring in text other people can write (demo_whatsapp_wrapped and get_conversation_snapshot); 1 tool can send data out of your system (generate_whatsapp_link). 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 3 findings (annotation-conflict and credentials-mention). 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 14, 2026, version 0.4.0.

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.

8 tools

Deletes or cancels 1Claude Code: deny
  • disconnect_conversationUse only when the user explicitly asks to disconnect or revoke this connection.1 finding

    Why it is in this group

    • The description says "cannot be restored".
    • The server marks it destructive (destructiveHint: true).

    Findings

    • Info

      The description mentions "credential".

      Matched text

      description: "...nt OAuth grant immediately; the revoked credential cannot be restored. Fresh explicit cons..."

      What this check looks for
    Description, as the server wrote it
    Use only when the user explicitly asks to disconnect or revoke this connection. Permanently revoke its current OAuth grant immediately; the revoked credential cannot be restored. Fresh explicit consent creates a new grant while the original session is eligible. Stops future reads without deleting the source session or results already in this conversation. Other connections can be revoked on the website. This tool changes access and is destructive access revocation; it does not retrieve statistics or answer period comparisons.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: true
    • openWorldHint: false
    • title: Disconnect your selected conversation
Changes things 3Claude Code: ask
  • format_whatsapp_textConvert the user's Markdown source to text ready to paste into WhatsApp.1 finding

    Why it is in this group

    • The description's first verb is "convert": a word for a change.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    Findings

    • Warn

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

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The description's first verb is "convert": a word for a change.

      What this check looks for
    Description, as the server wrote it
    Convert the user's Markdown source to text ready to paste into WhatsApp. Before calling, copy the supplied source unchanged into text: do not Markdown-escape it or pre-convert it. Example: source **Meeting tomorrow** must be passed as **Meeting tomorrow**; the tool returns *Meeting tomorrow*. Preserve backslashes the user actually supplied, but never insert extra ones. After calling, display structuredContent.formattedText verbatim in a plain-text fenced code block, without escaping or reformatting it. Supports # headings, bold, italics, strikethrough, lists and links. Preserves URL destinations, literal text and fenced/inline code. Unsupported blocks remain literal; does not render HTML, summarize or translate. Inputs are processed by this server to return the result; this tool does not store or log phone numbers or text. It does not access an account or send messages.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
    • title: Format text for WhatsApp
  • style_whatsapp_textRewrite the user's text in one decorative Unicode style to paste into WhatsApp.No 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.
    Description, as the server wrote it
    Rewrite the user's text in one decorative Unicode style to paste into WhatsApp. Styles: bold, italic, monospace, bubble, gothic, double (double-struck), strikethrough, upside-down. Call it once the user has picked a style from this list. These are lookalike characters, not real fonts: accented letters and digits usually stay plain, and screen readers can read the result badly. After calling, display structuredContent.styledText verbatim in a plain-text fenced code block. Inputs are processed by this server to return the result; this tool does not store or log phone numbers or text. It does not access an account or send messages.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
    • title: Restyle text with decorative letters
Only reads 4Show 4 read-only tools
  • demo_whatsapp_wrappedShow a synthetic WhatsApp recap without an account, upload or private conversation access.Reads private dataReads text others writeNo findings

    Why it is in this group

    • The description's first verb is "show": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "show" and "conversation": it reads messages.
    • untrusted-inputThe description's first sentence says "show" and "conversation": it reads messages, which other people can write.
    Description, as the server wrote it
    Show a synthetic WhatsApp recap without an account, upload or private conversation access. Use only for a demo or example, never to answer questions about the user's real conversation. All numbers are fictional and fixed; no relationship diagnosis. Returns a text and structured recap, without image export.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
    • title: Show a fictional WhatsApp Wrapped recap
  • get_analysis_resultsUse only to retrieve or explain the semantic analysis of the user's explicitly authorized Pro or Advanced WhatsApp export analysis.No findings

    Why it is in this group

    • The name "get_analysis_results" says "get": a word for reading.
    Description, as the server wrote it
    Use only to retrieve or explain the semantic analysis of the user's explicitly authorized Pro or Advanced WhatsApp export analysis. Pro: sentiment, engagement, conflict, clarity and relationship health. Advanced, Client Conversation Audit: the same without relationship health, plus audit counts (commitments, questions, objections, risks, next-action priorities). Advanced, Relationship Deep Dive: all of Pro plus whole-conversation responsiveness numbers per participant and counts of phases, loops, conflict episodes, repairs, strengths and plan steps. Returns scores, levels, whole-conversation trends, per-participant labels and key-finding counts. Refer to it as the user's Pro or Advanced analysis (source.tier), not as paid or extended results. The scores and levels are Le Confidant's own results; explain them, do not extend them into diagnoses. For message counts and reply timing use get_conversation_snapshot instead. DO NOT call for date-window, month-to-month or period-comparison requests. Reads stored results only and never starts analysis; sections not finished on the website are listed as pending. Returns neutral A/B labels, never names, summaries, descriptions or excerpts. Free sessions get an error. Set locale to the user's supported language.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
    • title: Read the semantic analysis
  • get_conversation_snapshotUse only to retrieve or explain WHOLE-SESSION numerical statistics from the user's explicitly authorized WhatsApp export analysis.Reads private dataReads text others writeNo findings

    Why it is in this group

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

    What it can reach

    • private-dataThe name "get_conversation_snapshot" says "get" and "conversation": it reads messages.
    • untrusted-inputThe name "get_conversation_snapshot" says "get" and "conversation": it reads messages, which other people can write.
    Description, as the server wrote it
    Use only to retrieve or explain WHOLE-SESSION numerical statistics from the user's explicitly authorized WhatsApp export analysis. DO NOT call for date-window, month-to-month, trend or period-comparison requests, even to check available data. For example, 'Compare September and November in my own connected WhatsApp analysis' needs NO tool call: explain that period comparison is unavailable. This tool returns only stored whole-session totals and cannot answer that request. For supported requests, returns neutral A/B labels, counts/shares, starts, reply timing/sample sizes, mean message length, definitions, limitations and expiry from the OAuth-selected session. No raw messages, names, excerpts, paid insights or new analysis. Set locale to the user's supported language.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
    • title: Read whole-session statistics
  • normalize_whatsapp_numbersValidate up to 100 phone numbers and return E.164 numbers and wa.me links with per-row status and duplicate positions.No findings

    Why it is in this group

    • The description's first verb is "validate": a word for reading.
    Description, as the server wrote it
    Validate up to 100 phone numbers and return E.164 numbers and wa.me links with per-row status and duplicate positions. National numbers require defaultCountry. Never guess a country from language. Does not check WhatsApp registration or contact people. Inputs are processed by this server to return the result; this tool does not store or log phone numbers or text. It does not access an account or send messages.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
    • title: Normalize phone numbers for WhatsApp

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.

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

Reads private data2
Reads files, mail, messages, databases, secrets, calendars or repositories.
Reads text others write2
Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
Can send data out1
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__le-confidant__generate_whatsapp_link"
    ]
  }
}

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 16, 2026, 14:21 UTC, unchanged since Nov 2, 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
4
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
https://leconfidant.ai/.well-known/oauth-protected-resource/api/mcp
Authorization servers
https://leconfidant.ai
Legacy HTTP+SSE transport
no
Dynamic client registration
no
Client ID metadata documents
yes
  • HighThe whole server

    The endpoint served tools/list without credentials, and the list has 4 tools that change things (1 destructive, 3 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://leconfidant.ai/api/mcp: tools/list returned 8 tools to an anonymous client, including disconnect_conversation (destructive), generate_whatsapp_link (write), format_whatsapp_text (write), style_whatsapp_text (write)

    What this check looks for

Findings

Matches of fixed checks, with the text they matched.

2 warn1 info
  • The server marks "generate_whatsapp_link" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The name "generate_whatsapp_link" says "generate": a word for a change.

    What this check looks for
  • The server marks "format_whatsapp_text" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The description's first verb is "convert": a word for a change.

    What this check looks for
  • The description mentions "credential".

    Matched text

    description: "...nt OAuth grant immediately; the revoked credential cannot be restored. Fresh explicit cons..."

    What this check looks for

History

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

  1. Tools1 added.+1 addedView diff

Install

Pinned to version 0.4.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http le-confidant https://leconfidant.ai/api/mcp
claude mcp add --transport http le-confidant https://leconfidant.ai/api/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__le-confidant__normalize_whatsapp_numbers",
      "mcp__le-confidant__demo_whatsapp_wrapped",
      "mcp__le-confidant__get_conversation_snapshot",
      "mcp__le-confidant__get_analysis_results"
    ],
    "ask": [
      "mcp__le-confidant__generate_whatsapp_link",
      "mcp__le-confidant__format_whatsapp_text",
      "mcp__le-confidant__style_whatsapp_text"
    ],
    "deny": [
      "mcp__le-confidant__disconnect_conversation"
    ]
  }
}

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 le-confidant. Review it before you use it.

Capability tags behind 3 rules
  • mcp__le-confidant__generate_whatsapp_link outbound
  • mcp__le-confidant__demo_whatsapp_wrapped private-data, untrusted-input
  • mcp__le-confidant__get_conversation_snapshot private-data, 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/ai.leconfidant/le-confidant/verify?hash=9a07510c783b998538a8d7487d850555e856c4baf2b08f033f59640b54c1df12" | jq -e '.changed == false'

Pinned hash 9a07510c783b, from Nov 14, 2026, 14:17 UTC.

Watch

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

https://radar.anlyon.com/servers/ai.leconfidant/le-confidant/changes.xml

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

Badge

Anlyon Radar badge for ai.leconfidant/le-confidant
[![Anlyon Radar](https://radar.anlyon.com/badge/ai.leconfidant/le-confidant.svg)](https://radar.anlyon.com/servers/ai.leconfidant/le-confidant)

Source: MCP Registry. Synced Nov 1, 2026, 14:17 UTC.

Le Confidant MCP server: questions and answers

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

Is the Le Confidant MCP server safe?

No scan can promise that. What its definitions show: 8 tools, 4 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 Le Confidant MCP server delete data?

Yes. 1 tool can delete, cancel or revoke something: disconnect_conversation.

Can the Le Confidant 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 Le Confidant MCP server have?

8 tools: 1 destructive, 3 write, 4 read. disconnect_conversation, generate_whatsapp_link, format_whatsapp_text, style_whatsapp_text, normalize_whatsapp_numbers, demo_whatsapp_wrapped, get_conversation_snapshot and get_analysis_results.

Does the Le Confidant MCP server need sign-in or OAuth?

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

Does the Le Confidant MCP server send data to outside services?

1 tool can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): generate_whatsapp_link.

Has the Le Confidant MCP server changed its tools?

Yes, once since Anlyon Radar started tracking it, most recently on Nov 14, 2026. Every change is on this page and in its Atom feed.