Ai Guardrails
Apply AI guardrails from your agent via MCP so prompts, outputs, and tool use get policy checks before users see risky content.
Overview
io.github.lazymac2x/ai-guardrails is a MCP server for the Ship phase that exposes AI guardrail and policy tooling to your agent via Cloudflare Workers streamable HTTP.
What is this MCP server?
- Remote MCP at https://api.lazy-mac.com/ai-guardrails/mcp (streamable-http)
- Version 1.0.0 with source repo lazymac2x/ai-guardrails-api
- Cloudflare Workers deployment for edge-adjacent guardrail calls from agents
- Lets coding agents invoke guardrail checks as tools during implementation and pre-release review
- Stacks with ai-gateway and ai-model-router for safer end-to-end LLM traffic
- 1 remote MCP endpoint
- GitHub: lazymac2x/ai-guardrails-api
What problem does it solve?
Your agent product can emit unsafe or off-policy content, and you lack a simple MCP-native way to enforce guardrails without building moderation infrastructure yourself.
Who is it for?
Indie builders adding a dedicated guardrails layer to Claude Code or Cursor-driven workflows before shipping user-facing AI features.
Skip if: Regulated deployments that mandate fully self-hosted moderation, human-in-the-loop review only, or zero reliance on external guardrail APIs.
What do I get? / Deliverables
After registration, your agent can call ai-guardrails MCP tools so security-oriented checks sit alongside generation and integration steps.
- MCP-discoverable guardrail tools for your agent
- Repeatable pre-ship safety checks invoked from the IDE
- Composable safety layer alongside gateway and router MCP servers
Recommended MCP Servers
Journey fit
How it compares
Policy-oriented MCP security service, not a generic code linter skill or passive monitoring dashboard alone.
Common Questions / FAQ
Who is io.github.lazymac2x/ai-guardrails for?
Solo builders and small teams shipping agent-backed products who need MCP-accessible guardrails hosted on Cloudflare Workers.
When should I use io.github.lazymac2x/ai-guardrails?
Use it during Ship security hardening and while building agent-tooling whenever prompts, tools, or outputs need explicit policy validation.
How do I add io.github.lazymac2x/ai-guardrails to my agent?
Add remote MCP configuration for https://api.lazy-mac.com/ai-guardrails/mcp with streamable-http transport, then enable the exposed guardrail tools in your agent.