Vettly
Wire your agent to moderate user-generated text, images, and video before they reach production UIs or logs.
Overview
Vettly is a MCP server for the Ship phase that moderates text, images, and video with configurable policies for AI applications.
What is this MCP server?
- Moderates text, images, and video through one MCP integration
- Policy-driven thresholds you can tune per product surface
- Built for AI applications where model and user output mix
- Stdio MCP server for Claude Code, Cursor, and similar agents
- Reduces manual review load on solo builders shipping chat or media features
- Multimodal moderation: text, images, and video
- MCP server version 0.1.0 per server.schema.json
What problem does it solve?
Solo builders shipping AI features risk unsafe or policy-breaking content slipping through when moderation is manual or split across vendors.
Who is it for?
Indie devs adding UGC or multimodal outputs to an AI app who need one MCP-backed moderation layer before launch.
Skip if: Teams that only need spam filtering on plain email or who already run enterprise trust-and-safety with dedicated analysts.
What do I get? / Deliverables
After you register Vettly, your agent can enforce consistent moderation policies on multimodal content before it reaches users.
- Agent-callable moderation for text, images, and video
- Configurable policy enforcement at integration time
Recommended MCP Servers
Journey fit
How it compares
MCP moderation integration, not an agent skill that rewrites prompts for safety.
Common Questions / FAQ
Who is Vettly for?
Solo and small teams building AI apps that accept or generate text, images, or video and need programmatic policy checks via MCP.
When should I use Vettly?
Use it in the Ship security subphase when you are wiring production guardrails before exposing user-facing or logged model output.
How do I add Vettly to my agent?
Add the io.github.code-with-brian/vettly MCP server entry to your agent’s MCP config per the Model Context Protocol registry flow for that server.