growth.engineer Opens a GTM Catalog Built for AI Agents

Your AI agent has a pile of tools and still grabs the wrong one. The missing piece is not another integration — it is knowing which tool to use, when, and what should happen next. growth.engineer is an open-source catalog built to give agents that missing layer for go-to-market work.

It is free to read, it speaks Markdown and MCP, and it is written so an agent can search it instead of a person clicking around.

What Happened

growth.engineer is an open-source catalog of GTM companies, tools, and workflows, published in a form AI agents can discover and use through Markdown and MCP. The catalog is organized into a few kinds of entries:

• Company entries that list each company's MCP, CLI, and API options, plus the credentials it requires.
• Tool-level files covering MCP tools, CLI commands, and API endpoints.
• Multi-step workflows that chain several tools together, up to 10 steps.
• Read-only MCP, so an agent can search and retrieve catalog data without changing anything.
• Action checks that require approval before an agent sends a message, spends money, or changes data.

The post frames the point simply: the failure was never a shortage of tools. It was that agents do not know which one fits, when to reach for it, or what follows.

Why It Matters

The catalog is aimed at the gap between an AI that has tools and an AI that uses the right tool at the right moment. Knowledge plus tools is the old recipe. Knowledge, tools, a workflow, and guardrails is the one being pushed here.

Those guardrails are the part worth noticing. An agent that can send messages or spend money on its own is a risk you inherit, not a feature you enjoy, and building the approval step into the catalog is a quiet admission that the step belongs there from the start. If your team is wiring agents into go-to-market work, this is a catalog to read before you write another integration yourself.