Deliver Prompts Into the Tools Where Work Actually Happens

Keep one governed source of truth for your prompts, then pull the right version — plus its context — into agents on n8n, Make, MCP, LangChain, CrewAI, and your own code, at runtime.

Connect Your Agents Read the Guides

Runtime Destinations

n8n

Community node

Make.com

HTTP module

Zapier

Action step

MCP

MCP server

LangChain

Prompt loader

CrewAI

Tool adapter

SDKs

TypeScript & Python

Your code

REST API

PromptPal Runtime

One Prompt Source of Truth, Delivered Everywhere

Stop pasting prompts into scripts and workflow nodes. Resolve them from PromptPal at runtime so every agent uses the approved version — with the context it needs.

No-Code Automations

Drop PromptPal into n8n, Make, or Zapier and resolve the right prompt in a workflow step — no redeploys when the prompt changes.

Agent Frameworks

First-class loaders for LangChain and CrewAI, plus an MCP server so any MCP-native agent can search, get, and resolve prompts.

SDKs & REST API

TypeScript and Python SDKs over a simple REST delivery API. `resolve` and `assemble` endpoints return runtime-ready prompts and context.

Governed Context, Not Just Prompts

Agents pull the prompt and the governed facts, policies, and procedures from your Brain in one resolve — so answers stay grounded.

Version-Pinned Delivery

Pin a prompt version per environment or always take the approved latest. Ship a prompt change once and every runtime picks it up.

Run Attribution & Analytics

Every connector logs runs back to PromptPal, so you can see which prompt and context versions actually performed — across every platform.

Two Ways to Connect — Same Source of Truth

Use MCP where your agent supports it, or the REST API anywhere else. Both resolve from the same governed prompts and context.

MCP rail

LangChain · CrewAI · Claude Desktop · Cursor · n8n MCP node

For MCP-native clients, PromptPal runs as an MCP server exposing prompt search, get, resolve, and assemble as tools your agent can call directly.

REST rail

Make · Zapier · custom apps · any HTTP node

Everything with an HTTP request can call the same delivery API. `POST /resolve` and `/assemble` return the prompt plus governed context, ready to run.

Agent Platforms FAQ

Which platforms can I deliver prompts to?+

No-code tools like n8n, Make, and Zapier; agent frameworks like LangChain and CrewAI; any MCP-native client; and your own code via the TypeScript and Python SDKs or the REST API.

What is the difference between MCP and the REST API?+

MCP lets MCP-native agents call PromptPal as a set of tools. The REST API is the universal path — any platform that can make an HTTP request can resolve a prompt. Both return the same governed prompt and context.

Do agents get context too, or just the prompt?+

Both. A single resolve returns the prompt plus the governed facts, policies, and procedures from your Personal or Company Brain, so agents stay grounded in your approved knowledge.

What happens when I update a prompt?+

You publish the change once in PromptPal. Every connected runtime picks up the approved version on its next resolve — no redeploys and no copy-paste drift across workflows.

Can I see how prompts perform in each platform?+

Yes. Connectors log runs back to PromptPal, so analytics show which prompt and context versions actually ran and performed — across every destination.

Ship One Prompt. Run It Everywhere.

Connect PromptPal to your agents and workflows, and deliver governed prompts and context into every runtime you use.

Get Started Free See Setup Guides