Cedric MCP server
OfficialLets Cedric act as an AI employee within Discord team chats, supporting communication and task handoff in the places the team already talks.
Connects Cedric to GitHub so agents can work with repositories, pull requests, and related development workflows through connected-app actions.
Enables searching and sending email through Gmail, with approval safeguards before messages are sent.
Enables checking calendar free/busy availability and creating calendar events, with approval safeguards for calendar changes.
Connects Cedric to HubSpot so agents can work with CRM data and run connected-app actions or recipes through Cedric.
Connects Cedric to Linear so agents can work with issues and project tracking, including opening or managing Linear issues through connected-app actions.
Connects Cedric to Notion so agents can access and work with workspace documents, knowledge, and connected-app content.
Lets Cedric operate in Slack team channels, including reading channel context, sending messages and DMs, posting votes, and coordinating approvals for team workflows.
Connects Cedric to Stripe so agents can work with payment and billing operations through connected-app actions or recipes.
Lets Cedric act as an AI employee within Telegram team chats, supporting communication and task handoff in the places the team already talks.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Cedric MCP serverwhat did we decide about the new pricing plan?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Cedric is an AI employee for Slack, Telegram and Discord. He joins your meetings, works in the tools you connect (HubSpot, Notion, Gmail, Google Calendar, Linear, GitHub, Stripe and 3,000+ more) and asks before anything is sent or changed.
This repo is how you reach him from code. Cedric is a remote MCP server, so Claude Code, Cursor, VS Code, Claude, ChatGPT, Codex CLI, Gemini CLI, Windsurf or your own agent can use his 129 tools: team memory, channels, connected apps, meetings, files, workflows and scheduling. Your agent is the brain; Cedric is the hands.
Endpoint |
|
Transport | Streamable HTTP |
Auth | Either way works: sign in once in the browser (OAuth 2.1 with PKCE, nothing to paste), or send an API key |
Account | Free to start, no card: app.meet-cedric.com |
Quick start
Claude Code:
claude mcp add --transport http cedric https://app.meet-cedric.com/api/v1/mcp
# then run /mcp, pick cedric, and choose AuthenticateCursor (~/.cursor/mcp.json):
{
"mcpServers": {
"cedric": { "url": "https://app.meet-cedric.com/api/v1/mcp" }
}
}VS Code:
code --add-mcp '{"name":"cedric","type":"http","url":"https://app.meet-cedric.com/api/v1/mcp"}'Claude, ChatGPT, Codex CLI, Gemini CLI, Windsurf and API keys: see docs/clients.md.
Once connected, call the getting_started tool. It returns the full guide, including when to act directly and when to hand a whole task to Cedric with ask_cedric.
Related MCP server: StateMCP
What your agent can do with Cedric
A few of the 129 tools:
Area | Tools |
Hand over a whole task |
|
Team memory |
|
Chat |
|
Connected apps |
|
Email and calendar |
|
Meetings |
|
Schedules and workflows |
|
Files and documents |
|
Most tools run as soon as you call them. Under a browser sign-in or a standard key, two kinds don't:
Direct key only: sending mail or DMs, changing the calendar, calling a connected app's API (
integration_request,run_recipe), running code and compiled automations (schedule_automation) are refused withrequires_direct_key. Use a direct key, or hand the task toask_cedric, which asks a person before anything goes out.Needs approval: posting in a channel, moving an event hook and starting a build return
pending_approvaluntil a person approves.
See approvals.
What people build with it
"What did we decide?" from the editor. Before touching the pricing code, your coding agent asks
recall_memoriesand gets the team's actual decision, not a guess.Ship and tell. When a PR merges, the agent writes the summary,
send_channel_messageposts it in #eng once a teammate approves, andrememberkeeps the decision for next time.CI that speaks up. A failed deploy calls
/api/v1/chat: "Tell #eng what broke and open a Linear issue." Cedric drafts both and a teammate approves with one tap.Settle it with a vote.
post_voteasks the channel which name to ship, andvote_resultsbrings the tally back to your agent.
More, with the calls spelled out: docs/use-cases.md.
Use it from code
The official MCP SDKs connect with an API key from Cedric's Settings → MCP.
TypeScript (examples/node/mcp-client.mjs):
import { Client } from "@modelcontextprotocol/sdk/client/index.js";
import { StreamableHTTPClientTransport } from "@modelcontextprotocol/sdk/client/streamableHttp.js";
const transport = new StreamableHTTPClientTransport(
new URL("https://app.meet-cedric.com/api/v1/mcp"),
{ requestInit: { headers: { Authorization: `Bearer ${process.env.CEDRIC_KEY}` } } },
);
const client = new Client({ name: "my-agent", version: "1.0.0" });
await client.connect(transport);
const { tools } = await client.listTools();
const result = await client.callTool({
name: "recall_memories",
arguments: { query: "pricing decisions" },
});Python (examples/python/mcp_client.py):
from mcp import ClientSession
from mcp.client.streamable_http import create_mcp_http_client, streamable_http_client
http = create_mcp_http_client(headers={"Authorization": f"Bearer {key}"})
async with http, streamable_http_client(URL, http_client=http) as (read, write):
async with ClientSession(read, write) as session:
await session.initialize()
result = await session.call_tool("recall_memories", {"query": "pricing decisions"})No agent of your own? One HTTP call with an API key runs a whole Cedric turn (examples/node/chat.mjs, examples/python/chat.py):
curl -X POST https://app.meet-cedric.com/api/v1/chat \
-H "Authorization: Bearer $CEDRIC_KEY" \
-H "Content-Type: application/json" \
-d '{"message": "Summarize this week for the team"}'
# { "reply": "...", "session": "abc123", "outcome": "answered" }Chat sessions, file uploads, approvals and limits: docs/api.md.
In this repo
docs/clients.md setup for every MCP client, API keys, how the sign-in works
docs/tools.md every tool, grouped, with its parameters
docs/use-cases.md worked examples of what to build with Cedric
docs/api.md MCP, chat and files endpoints, approvals, limits
examples/node/ MCP SDK client and chat call (TypeScript SDK)
examples/python/ MCP SDK client and chat call (Python SDK)
examples/curl/cedric.sh the whole API from the shell
server.json the MCP Registry entryLinks
Website: meet-cedric.com
Docs: meet-cedric.com/docs
Connect other MCP servers to Cedric: bring your own tools
Pricing: meet-cedric.com/pricing
Security: meet-cedric.com/security
Contact: sales@meet-cedric.com
Cedric is built in Zurich, Switzerland, by SFF Studio. The examples and docs in this repo are MIT licensed; the Cedric service itself is not open source.
This server cannot be deployed
Maintenance
Related MCP Connectors
MCP server for your apps' tools and custom tools, plus hosted AI agents and approval-gated workflows
Connect any AI agent to 1,000+ apps and 27,000+ actions through one remote MCP server (OAuth).
Remote streamable-HTTP MCP server running on a single Cloudflare Worker. Your assistant gets live Airbnb, Amazon, Booking.com, Google Flights, Maps and Reddit data, social search on X, Instagram and TikTok, the Meta Ad Library, and image/video generation without any keys. Connect your own accounts to let it send WhatsApp or Telegram messages, work an IMAP inbox, manage Meta Ads campaigns and publish to X and LinkedIn. OAuth 2.1 with PKCE; stored credentials are AES-256-GCM encrypted.
Zero-setup MCP gateway securely connecting AI to your tools with authentication and workflows
Related MCP Servers
- AlicenseNot gradedqualityBmaintenanceMCP server for Microsoft Teams that lets AI assistants search, message, and manage chats, meetings, files, and contacts using your existing Teams browser login—no app approval or Azure AD registration needed.1,327 npm1MIT
- AlicenseNot gradedqualityBmaintenanceLets you expose any product, web app, or SaaS backend as a dynamic, stateful MCP server with built-in agent memory, context persistence, and workflow-aware tool gating for AI agents.0MIT
- AlicenseNot gradedqualityCmaintenanceEnables MCP clients to discover and create agents, durably submit messages with verifiable replies, and search source-linked local memory. Exposes eight authenticated tools with owner-approved OAuth, transcript-backed verification, and explicit control over history import and embedding.MIT
- AlicenseNot gradedqualityBmaintenanceExposes a remote Streamable HTTP MCP endpoint that lets an AI client save, search, and retrieve structured memories, archived conversations, social-account notes, and knowledge documents. Access is scoped by per-client API keys with configurable read/write permission levels, collections, and operations.ISC