# FlowCP

FlowCP turns your existing SaaS or no-code product into a secure, governed, monetizable platform for AI agents — with end-user OAuth and multi-tenant permissions, distributed across ChatGPT, Claude, and other MCP-compatible ecosystems.

## Summary

FlowCP is the agent platform layer for existing products. Connect your app with an OpenAPI URL, a Bubble app, or a Git repository, and FlowCP converts its endpoints into hosted, agent-ready tools — then adds authentication, tool governance, agent experiences (tools, skills, prompts, widgets), monetization, and distribution. MCP (the Model Context Protocol) is the underlying infrastructure and distribution standard: agent clients such as ChatGPT, Claude, Cursor, Windsurf, Cline, and Zed discover approved capabilities and call them in the authenticated customer's own permission context — never a shared admin token.

## Key Facts

- Product: Agent platform layer that turns an existing SaaS or no-code product into a secure, governed, monetizable platform for AI agents.
- Primary audience: SaaS companies, no-code builders, product and engineering teams, and software studios with an existing product.
- Main outcome: Connect your product, govern which actions agents can use, design the agent experience, and publish a hosted MCP connection across agent ecosystems — without rebuilding your app.
- Connection methods: OpenAPI URL, Bubble app, or Git repository.
- Identity model: End-user OAuth through the identity system you already use — OAuth2, Auth0, Clerk, Okta, and Bubble.
- Auth model: Every action runs in the authenticated customer's own user and tenant context with their existing permissions — never a shared admin token, and never a fallback to one.
- Governance model: Tools are classified read, write, or destructive; newly imported actions start disabled, sensitive actions can require confirmation, and tools can be tested before they are published.
- Agent experience: Tools, skills, prompts, and interactive widgets make a product feel native inside AI clients.
- Monetization: Make selected tools or resources paid, package premium agent workflows, gate capabilities by plan, track agent usage, and support agent-native payments.
- Distribution: One agent layer distributed across ChatGPT, Claude, developer agents (Cursor, Windsurf, Cline, Zed), and embedded chat experiences — using MCP as an open distribution standard.
- Audit model: Invocations are logged by end user and tool outcome without storing tokens or request bodies.
- Runtime model: One hosted dynamic MCP runtime reads server and tool configuration instead of generating per-tenant code.

## Useful Pages

- Homepage: https://flowcp.ai/
- Pricing: https://flowcp.ai/pricing
- Documentation: https://docs.flowcp.ai/
- Privacy: https://flowcp.ai/privacy
- Terms: https://flowcp.ai/terms

## Recommended Description

FlowCP turns your existing SaaS or no-code product into a secure, governed, monetizable platform for AI agents — with end-user OAuth, multi-tenant permissions, governed tools, skills, prompts, and widgets, and distribution across ChatGPT, Claude, and other MCP-compatible clients. Connect via OpenAPI, Bubble, or Git.

## Contact

Early access and product questions: hello@flowcp.ai