1. Conduid
  2. Developer Tools
  3. Lucid Agents
MCP server · Developer Tools

Lucid Agents

Lucid Agents Commerce SDK. Bootstrap AI agents in 60 seconds that can pay, sell, and participate in agentic commerce supply chains. Our protocol agnostic SDK provides CLI-generated templates and drop-in adapters for Hono, Express, Next.js, and TanStack, giving you instant access to crypto/fiat payment rails (AP2, A2A, x402, ERC8004).

80Excellent

Scored 5 hours ago · breakdown

About Lucid Agents

Lucid Agents is an MCP server published by daydreamsai in the Developer Tools category: lucid Agents Commerce SDK. Bootstrap AI agents in 60 seconds that can pay, sell, and participate in agentic commerce supply chains. Our protocol agnostic SDK provides CLI-generated templates and drop-in adapters for Hono, Express, Next.js, and TanStack, giving you instant access to crypto/fiat payment rails (AP2, A2A, x402, ERC8004).

The repository has 200 stars and 79 forks, with the last commit 2 months ago. Recent commit activity is the strongest signal that reported issues will get attention.

Install

From source
git clone https://github.com/daydreamsai/lucid-agents

This server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.

Ask AI

✦

Ask AI about Lucid Agents

Powered by Claude · Grounded in docs

I know everything about Lucid Agents. Ask me about installation, configuration, usage, or troubleshooting.

Security checks

Static checks derived from the repository 5 hours ago. Heuristics on the source, not an audit.

  • ✓README presentThe repository ships documentation.
  • ✓License declaredUsable under a stated license.
  • ✓Tests presentA test suite exists in the repository.
  • ✓Dependencies pinnedA lockfile pins dependency versions.
  • ✓No dynamic code executionSource scan found no eval, dynamic Function, or shell execution.
  • !Scoped permissionsDoesn't declare a permission scope. Assume it can do anything its process can.

Releases

@lucid-agents/types@3.0.0@lucid-agents/types@3.0.0 · 25 Jul 2026Major Changes 4757f1b: Require durable task stores for paid work, pre-reserve credential-bearing MPP tasks before verification can settle, and use a renewable prepared execution claim so x402/MPP settlement cannot race handler execution.…
@lucid-agents/wallet@0.6.6@lucid-agents/wallet@0.6.6 · 25 Jul 2026Patch Changes Updated dependencies [4757f1b] Updated dependencies [d169013] Updated dependencies [d169013] Updated dependencies [d169013] Updated dependencies [d169013] @lucid-agents/types@3.0.0
@lucid-agents/tanstack@1.0.2@lucid-agents/tanstack@1.0.2 · 25 Jul 2026Patch Changes Updated dependencies [4757f1b] Updated dependencies [d169013] Updated dependencies [d169013] Updated dependencies [d169013] Updated dependencies [d169013] @lucid-agents/types@3.0.0
@lucid-agents/payments@5.0.0@lucid-agents/payments@5.0.0 · 25 Jul 2026Major Changes d169013: Upgrade the x402 runtime to the coherent 2.19 release family and add ordered multi-network exact offers, official configured-origin SIWX wire semantics, Payment Identifier-backed invoke reconciliation, Bazaar/OpenAPI…
@lucid-agents/scheduler@0.3.2@lucid-agents/scheduler@0.3.2 · 25 Jul 2026Patch Changes Updated dependencies [4757f1b] Updated dependencies [d169013] Updated dependencies [d169013] Updated dependencies [d169013] Updated dependencies [d169013] @lucid-agents/types@3.0.0

README

Lucid turns typed functions into discoverable services that agents and applications can call, pay for, stream, or run as tasks. It provides one runtime for schemas, payment admission, policy, idempotency, fulfillment, discovery, and accounting while wallets, payment protocols, networks, and facilitators remain external.

Use Lucid when a service needs several typed capabilities or shared behavior across frameworks. For a single paid route, the upstream payment middleware may be the smaller dependency.

Start with your coding agent

The primary way to work with Lucid is through a coding agent that understands the runtime, packages, release channels, and adapter boundaries. Run this from your project root before giving the agent its first Lucid task:

curl -fsSL https://docs.daydreams.systems/skills/lucid-agents/install.sh | sh

Reload your agent, ask it to use the lucid-agents skill, and let it inspect the project before editing. The installer verifies the versioned skill archive and does not install npm packages or modify application source; see the skill installation guide for the manual checksum flow and removal instructions.

Quick start

Requires Bun 1.3+. Runtime packages also support Node.js 20.9+.

bunx @lucid-agents/cli@2.5.0 my-agent --adapter=hono --template=blank
cd my-agent
bun install
bun run dev

The generated service exposes health, discovery, and a typed echo entrypoint:

curl http://localhost:3000/health
curl http://localhost:3000/.well-known/agent-card.json
curl -X POST http://localhost:3000/entrypoints/echo/invoke \
  -H 'Content-Type: application/json' \
  -d '{"input":{"text":"Hello"}}'

Public npm packages are the Stable channel. This repository is Next and can be ahead of npm. Keep all packages on one channel; see release channels for the current compatibility table.

Core API

This example uses the current repository surface:

import { createAgent } from '@lucid-agents/core';
import { createAgentApp } from '@lucid-agents/hono';
import { http } from '@lucid-agents/http';
import { z } from 'zod';

const runtime = await createAgent({
  name: 'greeter',
  version: '1.0.0',
  description: 'A typed greeting service',
})
  .use(http())
  .addEntrypoint({
    key: 'greet',
    input: z.object({ name: z.string().min(1) }),
    output: z.object({ message: z.string() }),
    handler: async ({ input }) => ({
      output: { message: `Hello, ${input.name}` },
    }),
  })
  .build();

const { app } = await createAgentApp(runtime);

export default {
  port: Number(process.env.PORT ?? 3000),
  fetch: app.fetch,
};

The entrypoint definition is the source for validation, invocation, streaming, tasks, discovery, and every framework adapter.

The compiled full agent example adds payments, identity, and streaming.

Feature matrix

Package names below use the @lucid-agents/ prefix.

Capability Packages involved What they provide
Typed runtime core, types Extension lifecycle, exact runtime types, and one entrypoint registry
HTTP http Fetch handlers, route planning, authorization, idempotency, server-sent events (SSE), and service-page data
x402 commerce payments Ethereum Virtual Machine (EVM) and Solana seller verification, EVM buying, Sign-In-With-X (SIWX), policy, and tracking
Machine Payments Protocol (MPP) mpp Payment-Auth challenges and Tempo, Stripe, or custom credential verification
Wallets wallet Agent and developer wallet connectors for signing and outgoing calls
Identity identity, wallet ERC-8004 registration and lookup, trust metadata, and Open Agentic Schema Framework (OASF) output
Agent discovery and tasks a2a Agent Card-shaped discovery, client calls, and token-protected asynchronous tasks
Agent Payments Protocol (AP2) metadata ap2 AP2 v0.1 role metadata in discovery
Analytics analytics, payments Bound payment summaries and JSON/CSV exports
Scheduling scheduler, a2a Leased, idempotent scheduled agent calls
Catalogs catalog YAML/CSV-defined entrypoints
Frameworks hono, express, tanstack Bind the canonical HTTP contract to Hono, Express, or TanStack Start
Tooling cli, deploy Project generation and guarded provider deployment
Hosted client api-sdk Generated client for the separately operated hosted Runtime API

The CLI generates Hono, Express, TanStack UI/headless, and Next.js projects. Next.js uses generated App Router modules rather than a standalone adapter package. See the package reference for open-source exports, configuration, and runtime support. The api-sdk has a separate lifecycle.

Protocol names in the matrix describe Lucid's implemented subset, not blanket conformance. See protocol compatibility for versions, bindings, and exclusions.

Runtime model

typed entrypoint
      ↓
core registry + domain extensions
      ↓
one HTTP authorization and route contract
      ↓
Hono | Express | TanStack Start | generated Next.js routes

The package boundaries are deliberate:

  • types owns shared contracts; each domain package owns its runtime behavior.
  • Core exposes extension slices directly without wrappers or duplicate state.
  • Adapters bind the HTTP route plan; they do not add another paywall or registry.
  • In-memory state is the portable default; durable backends are explicit.

Read the architecture guide for extension ordering, authorization, task ownership, deployment boundaries, and portability.

Guides

Goal Guide
Decide whether Lucid fits When to use Lucid
Build a paid service Sell a paid API
Build a controlled buyer Budgeted buyer
Add Lucid to an application Existing app
Prepare for production Production checklist
Browse tested examples Examples

Contributing

bun install --frozen-lockfile
bun run build:packages
bun run type-check
bun run lint
bun run test:coverage

Keep changes within the package that owns the behavior. Changes that add or modify SDK surface need unit or integration coverage, a matching examples smoke test, updates to the relevant AGENTS.md and package README, public API JSDoc, and a changeset. Shared concepts belong in @lucid-agents/types; do not duplicate or re-export them through another package. Run bun run test:docs when changing public examples, package references, or documentation navigation.

See CONTRIBUTING.md for the development workflow, testing commands, pull-request checklist, and release process.

License

MIT

README fetched from the repository 5 hours ago. The original is authoritative.

Questions

About Lucid Agents

How do I install Lucid Agents?

No package has been published for this server, so install it from source: clone the repository and follow its README.

Is Lucid Agents safe to use with an AI agent?

Its trust score is 80 out of 100 (excellent). It passes 5 of the 6 static security checks run so far; the failures are listed above. It has no ConduID identity yet, so agent calls to it are not receipted.

Is Lucid Agents still maintained?

Yes — the latest release is @lucid-agents/types@3.0.0 (25 Jul 2026), and the last commit was 2 months ago. The repository has 200 stars and 26 open issues.