# Drive a full outbound department from your own stack. URL: https://zevari.ai/for/gtm-engineers > Drive Zevari's six AI SDR specialists from Claude Code, Codex, ChatGPT or your own code. Hosted LinkedIn MCP plus REST API, and every send waits for your approval until you switch a trained specialist to Autopilot. $87/mo. For GTM engineers and technical sales teams Zevari is an AI SDR team for LinkedIn and email: six specialists that find prospects, score them, write in your voice, follow up and qualify replies. GTM engineers drive the same specialists from Claude Code, Codex, ChatGPT or any MCP client over the hosted LinkedIn MCP, or from their own code over the REST API. State is hosted, and every send waits for your approval until you switch a trained specialist to Autopilot. Connect your agent [REST API reference](https://docs.zevari.ai/api/reference) Who this page is for GTM engineers, AI engineers, sales ops and the SDR leads who already work in Claude Code or Codex, and want outbound wired into the tools they use every day. This page replaces the separate pages we used to run for SDRs, GTM operators, AI engineers and Claude or Codex power users. If you would rather tell Zevari who you sell to and review the work in the app, start with [what an AI SDR team does](https://zevari.ai/ai-sdr) or [Zevari for founders](https://zevari.ai/for/founders). Building on your own agent runtime? See [LinkedIn for AI agents](https://zevari.ai/linkedin-for-ai-agents). What your agent drives ## Six specialists, not a pile of endpoints Your agent does not have to orchestrate LinkedIn from scratch. It works with the same department the app runs, and each specialist keeps its own hosted state. ### Prospecting Runs saved Sales Navigator searches and lands net-new people on your lists. ### Signals Finds LinkedIn posts by keyword and captures the people engaging with them. ### ICP Scoring Scores every lead against your ICP as a Strong fit, Moderate fit or Weak fit, with a written reason you can correct. ### Voice Writer Drafts connection notes and messages in your voice, trained on what you have sent. ### Follow-up Runs multi-step sequences on LinkedIn and email inside your working hours. ### Reply Qualifier Assesses replies and surfaces the qualified conversations worth your time. Every capability, grouped by specialist: [what the LinkedIn MCP can do](https://zevari.ai/linkedin-mcp/tools). The outcome it reports is the [qualified conversation](https://zevari.ai/qualified-conversation), not a reply count. Connect ## Three ways in, one approval queue Hosted HTTP transport. Nothing to install, no Docker, no Chrome profile. ### From an MCP client Claude Code, Codex, ChatGPT, or any MCP client. One hosted HTTP endpoint. Registering it does not sign you in: run /mcp in Claude Code, or codex mcp login in Codex, and pick the workspace and LinkedIn account in the browser. ``` claude mcp add --transport http zevari https://mcp.zevari.ai/mcp # then run /mcp in Claude Code, pick zevari and sign in # Codex codex mcp add zevari --url https://mcp.zevari.ai/mcp codex mcp login zevari ``` ### Committed to the repo Put the server in your project .mcp.json so everyone on the team connects to the same department on clone. ``` { "mcpServers": { "zevari": { "type": "http", "url": "https://mcp.zevari.ai/mcp" } } } ``` ### From your own code Workers, cron jobs, CRMs, or a custom agent. A bearer token over plain HTTP, the same hosted state and approvals as MCP. ``` # Same specialists over the REST API. Keys start with zvr_api_ curl https://api.zevari.ai/v1/contentTemplates/list \ -H "Authorization: Bearer $ZEVARI_API_KEY" ``` [Setup guide](https://zevari.ai/quickstart) [MCP reference](https://docs.zevari.ai/mcp/reference) [About the LinkedIn MCP](https://zevari.ai/linkedin-mcp) Why not point a raw agent at LinkedIn ## An agent can research. It cannot hold a campaign. Claude or Codex can research a prospect well, but it has no safe connection to LinkedIn and no place to keep a multi-step sequence between sessions. Zevari provides both. Campaigns, schedules, working-hours queues and approvals live on hosted infrastructure, so the sequence keeps advancing when your laptop is closed, and nothing reaches an account until someone approves it or switches a trained specialist to Autopilot. ## The limits your agent cannot talk its way past - Approval while you train: connection requests, messages, comments and posts are staged for your approval before they reach the account, until you switch a trained specialist to Autopilot. - A hosted connector: no browser extension, no cookie injection, nothing running on your machine. - Weekly connection-request ceilings of 100 on Free*, 150 on Premium, 200 on Sales Navigator. - * Free accounts: LinkedIn caps connection requests that carry a note at 3 a month. The weekly ceiling covers requests sent without a note, and Premium and Sales Navigator accounts get a far larger allowance for requests with a note. Most outreach uses a note, so Premium or Sales Navigator is the practical choice for real outbound - the full explanation is on /safety. - A daily cap on all actions (400 Free, 1,000 Premium, 2,000 Sales Navigator), burst limits and working hours, enforced server-side. [Read the safety model](https://zevari.ai/safety) FAQ ## Questions from technical teams ### Which AI clients can drive Zevari? Claude Code, Claude, Codex, ChatGPT and any other MCP client over the hosted LinkedIn MCP. Self-hosted runtimes such as OpenClaw and Hermes Agent, or your own code, use the REST API with a bearer token. ### MCP or the REST API? Use whichever your stack speaks. MCP is the fastest path when an AI client is doing the reasoning. The REST API drops into pipelines, cron jobs and workers. Both reach the same six specialists, the same hosted state and the same approval queue, across 191 MCP tools and 109 REST endpoints. ### What does Zevari add that my agent cannot do alone? Two things. A safe connection to LinkedIn and email that an unaided agent does not have, and hosted state: campaigns, schedules, working-hours queues and approvals keep advancing after your session closes. ### Can anything send without a human? Yes, once you switch a specialist to Autopilot, and not before. First you train your team: you teach ICP Scoring who your ideal customer is by reviewing the leads it scores, and Voice Writer drafts in your voice from a Voice DNA learned from your own writing, sharpened by every message you approve or edit. While you train them, every connection request, message, comment and post waits for your approval. When a specialist has earned it, such as Follow-up after ten approved sends in 30 days, switch it to Autopilot on the AI team page and it runs on its own, in your voice and to your ideal customer. You can switch it back to Reviews with you at any time. ### Does it run email as well as LinkedIn? Yes. LinkedIn is the primary channel. Add a mailbox and the same sequence runs email steps under the same approvals and limits. ### What does Zevari measure? Qualified conversations: a buyer who replied with real interest to outreach Zevari sent, and kept talking. Not every reply, and not a booked meeting. The definition is public and versioned. ### What does it cost? One plan with all six specialists: $87 a month, or $497 a year (save 52%). Extra LinkedIn accounts are $37 a month, so one workspace per client or brand is straightforward. First-time accounts get a 7-day free trial, and there is a 30-day money-back guarantee after the first charge. ### Can I connect Zevari's MCP with an access token instead of an OAuth sign-in? Yes. Besides the browser sign-in, Zevari offers MCP access tokens, which you create under Settings → Developers. A token needs no OAuth sign-in, so it suits automations, agents and servers that cannot open a browser. Each token is locked to one LinkedIn sender account and to one workspace (or, if you confirm it, to all your workspaces), expires after 30, 90 or 365 days, and can be revoked at any time. Agencies and teams with several workspaces or senders make one token per client and sender, so an automation never has to switch workspace or sender mid-run. ## Put the department behind your agent $87 a month, or $497 a year (save 52%). Extra LinkedIn accounts $37 a month. 7-day free trial and a 30-day money-back guarantee. [See pricing](https://zevari.ai/pricing). ``` claude mcp add --transport http zevari https://mcp.zevari.ai/mcp # then run /mcp in Claude Code, pick zevari and sign in # Codex codex mcp add zevari --url https://mcp.zevari.ai/mcp codex mcp login zevari ``` [Start the free trial](https://zevari.ai/auth/signup?redirect=/app) [Connect your agent](https://zevari.ai/quickstart)