Skip to content

Private Alpha

Lucid is currently in private alpha. Access is by invitation only. Features, APIs, and pricing are subject to change. Request access to get started.

Lucid Developer Platform

Deploy AI. Prove it's safe. Let users verify for themselves.

Your users have no reason to trust your AI—and they shouldn't have to. Lucid gives them cryptographic proof that your agents are running exactly as promised: no data logging, no tampering, no shortcuts on safety guardrails.

How It Works

1. Deploy agents in one command

Ship models with built-in safety auditors. No infrastructure wrangling, no security theater—just working, verifiable AI.

pip install lucid-clilucid loginLogged in as [email protected]
lucid apply --model llama-3.1-8b --profile chat[+] Environment created: env-abc123def456

Connection URL: https://env-abc123def456.serverless.lucid.ai
Model: meta-llama/Llama-3.1-8B-Instruct
Auditors: injection, toxicity, sovereignty

[+] Environment ready!

2. Monitor and verify in real-time

Watch every guardrail fire. Catch problems before your users do. The Observer dashboard shows exactly what's happening inside your agents—backed by hardware attestation, not just logs.

Lucid Observer Dashboard

3. Give users proof, not promises

Every agent gets an AI Passport—a cryptographic certificate your users can verify independently. They can see exactly what model is running, what safety checks are active, and where their data lives.

Sovereign Chat with AI Passport
AI Passport Details

4. Treat agents as colleagues, not APIs

With Verifiable Agent Pods (VAP), each agent gets an email, handle, and AI passport -- managed like a team member. Share a Google Doc with the agent's email. Add it to a Slack channel. The agent gets context the same way humans do, while owner accountability, credential inheritance, approval flows, and auditor settings cascades enforce governance underneath.

Start Here

New to Lucid? Follow these steps to go from zero to a running, monitored agent:

  1. Install the CLI — pip install lucid-cli
  2. Authenticate — lucid login
  3. Deploy your first agent — 2-minute serverless quickstart
  4. Monitor in Observer — see guardrails fire in real time
  5. Build a custom auditor — your first safety guardrail
  6. Write Cedar policies — declarative rules for your auditors
  7. Go to production — go-live requirements

Not sure where to start? See the Documentation Guide for role-based learning paths.

Quickstarts

Verifiable Agent Pods (VAP)

  • VAP Guide: Agent identity, app catalog, console, auditor settings cascade, and self-hosted registries.
  • Credentials Guide: MCP-first credential model — ScopedCredentials, tool enforcement, three-tier cascade, and the MCP install flow.

Go Deeper

MCP & Workflows

  • MCP: Model Context Protocol for tool federation.
  • Workflows: Compose deployment types into multi-agent applications.

Setup & Installation

Deployment

Build Custom Auditors

Understand the Architecture

Reference

Compliance

Lucid provides cryptographic evidence for 36+ regulatory frameworks.

Help