Zapier Open Sources GTM Agent Skills for Cursor, Codex, and Claude Code.

Zapier has released a free GitHub repository of go-to-market agent skills and pre-built automations designed to work with AI coding environments like Cursor, Codex, and Claude Code. The collection, announced on LinkedIn by CEO Wade Foster, includes workflows built from real internal use cases across marketing, sales, RevOps, customer advocacy, and support.

The repository is built on the Zapier SDK combined with the Model Context Protocol (MCP), a framework that lets AI assistants interact with external systems. Users can install the agents directly into their AI coding tool and immediately access automations without building integrations from scratch.

The initial release includes five GTM automations:

• Daily lead audits that flag leads slipping through the cracks
• No-lead-left-behind rollup in daily, weekly, and standup formats
• Cross-CRM opportunity sync to keep deal data consistent
• Customer deck builder for quick presentation assembly
• PR and media inbox triage for automated PR response workflows

Each automation is designed to handle repetitive GTM tasks that typically require manual checks or custom scripting. The lead audit automation, for example, scans CRM records daily and surfaces leads that haven't been contacted or updated within defined timeframes.

Beyond the automations, the repository includes skill files, which are modular capabilities that agents can combine to execute tasks. These skills span the full GTM stack: lead enrichment, email sequence triggers, deal stage updates, customer health scoring, and support ticket routing.

The approach builds on Zapier's previous work with MCP, which lets AI tools interact with thousands of apps through natural language commands. By packaging these skills as open-source modules, Zapier is making it easier for developers and operators to build custom GTM agents without reinventing common workflows.

This release follows Zapier's earlier launch of GTM agent skills, which demonstrated how the company was using AI agents internally. Open sourcing the code gives other teams a starting point for similar implementations.

The repository is designed to work in environments where developers and business users already spend time. Installing the agents requires a single command in tools like Cursor or Claude Code, and the skills become available immediately for the AI assistant to use.

This differs from traditional automation platforms where users configure workflows in a separate interface. Here, the agent lives inside the coding environment, and users can ask it to execute tasks or build new automations using natural language. The approach lowers the barrier for technical teams who want to automate GTM operations without leaving their development tools.

Zapier's SDK handles the connection layer to external apps, so developers don't need to manage API authentication or rate limits. The MCP framework provides the structure for agents to understand which skills are available and how to chain them together.

The GitHub repository is available now. Teams can fork it, modify the skills, or add new automations that fit their workflows. Zapier has indicated this is the first release, with more skills and automations planned based on community feedback.