ZelesIA

Registre officiel Recherche et web

Turn the LinkedIn posts you save into a knowledge base your assistant can search and cite.

Installation

Ajoutez ceci à la configuration MCP de votre client (Claude Desktop : claude_desktop_config.json ; Cursor : .cursor/mcp.json ; VS Code et les autres utilisent la même forme « mcpServers »).

Remote (streamable-http)
Claude Code
claude mcp add --transport http zelesia 'https://mcp.zelesia.com/mcp'
Configuration JSON (Claude Desktop, Cursor, VS Code)
{
    "mcpServers": {
        "zelesia": {
            "type": "http",
            "url": "https://mcp.zelesia.com/mcp"
        }
    }
}

Documentation

Tirée du README du projet sur GitHub, actualisée chaque semaine. Les liens ouvrent GitHub.

Sommaire
  1. ZelesIA MCP Server
  2. Connect it
  3. What the assistant reads
  4. Tools
  5. Access and privacy
  6. Support

ZelesIA MCP Server

Turn the LinkedIn posts you save into a knowledge base your assistant can search and cite.

You save a post on LinkedIn with one click. ZelesIA summarises it, groups it by topic, and keeps it. Connect this MCP server and Claude — or ChatGPT, or Cursor — answers from what you saved, citing the original post.

This repository is the connection guide for the hosted server. There is nothing to install.

Connect it

You need a ZelesIA account with something saved. The server only ever reads your own library.

Claude (web and desktop)

Settings → Connectors → Add custom connector, and paste https://mcp.zelesia.com/mcp. A ZelesIA consent screen opens; approve it and the tools appear.

Claude Code
claude mcp add --transport http zelesia https://mcp.zelesia.com/mcp
Cursor, VS Code and other MCP clients

Point them at the same URL over Streamable HTTP:

{
  "mcpServers": {
    "zelesia": {
      "type": "http",
      "url": "https://mcp.zelesia.com/mcp"
    }
  }
}
ChatGPT

The server also publishes the search / fetch pair that ChatGPT's connectors expect, backed by the same library. Add it as a custom connector with the endpoint above.

What the assistant reads

Your own library: every post you saved, summarised, with its key points and its link back to LinkedIn.

Tools

ToolWhat it does
library_mapAn index of your library: how many posts, over what date range, which topics and headlines. Cheap — call it before searching, so the assistant doesn't guess at words you never saved.
search_librarySearches your saved posts by meaning and by keyword. Returns the topic, a link to the original post, and an excerpt — usually enough to answer and cite without opening anything. Paginated by cursor.
search / fetchThe ChatGPT-compatible pair, over the same library.

Every tool is read-only. Nothing the assistant does can change or delete what you saved.

Access and privacy

  • OAuth 2.1. You approve the connection on a ZelesIA consent screen. No API key to paste, no password shared with the client.
  • Your library only. A grant reads the posts of the account that approved it, and nothing else.
  • Revocable. Remove the grant from your ZelesIA account and the connection stops working straight away.

Terms: zelesia.com/terms · Privacy, including how to reach us: zelesia.com/privacy

Support

Open an issue in this repository.

Questions fréquentes

Qu'est-ce que ZelesIA ?

Turn the LinkedIn posts you save into a knowledge base your assistant can search and cite.

Comment installer ZelesIA ?

Il fonctionne comme serveur distant. Dans Claude Code : claude mcp add --transport http zelesia 'https://mcp.zelesia.com/mcp'. Pour les autres clients, collez la configuration JSON de cette page dans leurs réglages MCP.

Faut-il une clé API pour ZelesIA ?

Son entrée au registre n'indique aucune clé ni secret obligatoire. Certaines fonctions peuvent tout de même demander un compte auprès du service auquel il se connecte.

ZelesIA est-il open source ?

Oui. Son code est sur GitHub sous licence MIT.

Qui maintient ZelesIA ?

Il est publié par ZelesIA-AI. AI.tn le référence sans le maintenir ni le recommander.

Issu du registre MCP officiel. Être listé n'est pas une recommandation : vérifiez le code et les permissions d'un serveur avant de lui donner accès à vos données.