PIPELINE FOR DEVELOPERS

Your agent does the thinking.
Your team sees the work.

Build lists, draft campaigns, and save the evidence with your own agent. Pipeline gives you the workspace, tools, and review steps.

claudeOpenAIgeminigrokUse your compatible MCP client or coding agent.

Start where you already work.

Pick a connection. The workspace and permissions stay the same.

Make your next campaign a command.

Use your coding agent to read context and prepare the work. Save the result in Pipeline so your team can review it.

Open CLI guide
CLI
npm install -g pipeline-gtm
pipeline init
pipeline agents setup --targets codex,claude-code --project .

A command ends in work you can review.

Keep lists, messages, and campaign drafts in the same product your team uses.

One connected agent team for LinkedIn outreach and content
Launchpad
⌘ KWorkspace owner
Research completeChoose the first campaign play
Founder-led sales teamsFounders and revenue leaders at B2B SaaS teams building repeatable LinkedIn outbound. 1,240 estimated peopleBest fit
Agencies scaling deliveryOutbound agencies replacing manual research and campaign setup. 860 estimated people
Revenue teams consolidating toolsTeams operating separate data, outreach, and inbox systems. 2,100 estimated people
Website, ICP, and positioning are ready
Illustrative product data

Watch one campaign move from company context to a meeting-ready LinkedIn reply.

Reason with your own agent.

Your agent reads context and skills, then writes the brief and drafts using its own subscription.

Save through typed tools.

Pipeline checks inputs and workspace access before saving the result.

Review before live actions.

Backend AI needs explicit consent. Sending and other external actions follow workspace approvals.

Build with Pipeline | Developer preview