Best for
- Avito sellers aiming to reduce manual operational time
- Advanced users needing to automate multi-store or multi-item management via APIs
A
Avito MCP Server
Community
Let an AI agent run your Avito account: reply to buyers, keep listings priced and promoted, and fulfill orders automatically.
GitHub source repository ↗
60FMRS · C
The server is feature-rich with a large toolbox, but its official status is unclear and requires trusting a large-scale proxy of 148 tools. Safety features like dual confirmation and dry-run are commendable, yet users must assess risks and configure carefully.
The Avito MCP Server connects your Avito account to AI agents, enabling them to automate chat replies, update listing prices and promotions, manage orders, upload shipping labels, and more, covering 148 tools across 18 official Avito APIs. The server includes robust safety features: money-spending or public-facing actions require a second confirmation, dry-run mode allows rehearsal, and idempotency keys prevent duplicate charges. It runs over stdio by default for local use and supports HTTP with OAuth 2.1 for remote access. Ideal for Avito sellers who want to reduce manual work, but requires careful permission management and understanding of security features.
Tools
messenger_get_chats
Retrieve chats with optional filters like chat_id, used to check for unread messages.
listing_update_price
Update item price, with dryRun option for rehearsal.
order_ship
Mark an order as shipped, supports idempotency key to avoid duplicate charges.
meta_confirm_action
Confirm a pending sensitive or financial action by providing the confirmation ID.
meta_health
Output server health information including version and counters.
meta_auth_status
Check authentication status, optional probe to force token refresh.
meta_capabilities
Report active mode, allow/deny counts, and feature flags.
Setup
- Obtain Client_id, Client_secret, and Profile_id from the Avito developer portal. 2. Add the above JSON to your MCP client configuration. 3. Restart your client and start using it.
claude_desktop_config.json
{
"mcpServers": {
"avito": {
"command": "npx",
"args": [
"-y",
"avito-mcp"
],
"env": {
"Client_id": "YOUR_CLIENT_ID",
"Client_secret": "YOUR_CLIENT_SECRET",
"Profile_id": "YOUR_PROFILE_ID"
}
}
}
}
Fit and risk
Not for
- Individuals without Avito accounts or developer credentials
- Developers requiring a sandbox environment for testing (Avito does not provide sandbox credentials)
- Users who need a fully offline solution without external dependencies
Required permissions
- Read and reply to chat messages (requires messenger permissions)
- Modify product listings, prices, and promotions (requires listings permissions)
- Manage orders and fulfill shipping (requires orders permissions)
- Access account balance and analytics data (requires financial permissions)
Risks and side effects
- All operations execute against the live production environment; mistakes may lead to financial loss or account issues.
- Credentials are stored in configuration; ensure secure storage to avoid leaks.
- By default, high-risk actions require confirmation, but misconfiguration could execute unintended operations.
- Reliance on external Avito API; subject to rate limits and service changes.
Troubleshooting
- 1. Ensure Node.js version is 22.12 or higher.
- 2. Check that environment variables (Client_id, Client_secret, Profile_id) are valid.
- 3. View logs: on macOS, see ~/Library/Logs/Claude/mcp-server-avito.log (macOS) for Claude Desktop.
- 4. Use the meta_health tool to check server status.
- 5. If connection fails, verify network and Avito API accessibility.
Use cases
Automatically handle customer inquiries and chat replies
Automate price updates and promotion management
Process orders and track shipping logistics
Fetch real-time data such as account balance and sales analytics
Supported clients
Claude DesktopFull support
Claude CodeFull support
CursorFull support
WindsurfFull support