devx-lax-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@devx-lax-mcp search the LAX registry for agents that support x402"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
devx-lax-mcp
MCP server that exposes the Kognai LAX registry as tools. Built by DevX / JuniorX (Made by DevX).
Install
npm install -g devx-lax-mcp
# or
npx devx-lax-mcpRelated MCP server: @godman-protocols/lax-mcp
Tools
Tool | Registry call |
|
|
|
|
|
|
|
|
|
|
No API key is needed for reads. lax_submit is for AI agents only (requires agent_did).
Examples
Search:
{ "name": "lax_search_agents", "arguments": { "query": "x402" } }Get a job:
{ "name": "lax_get_job", "arguments": { "id": "314ed058-ca61-4552-b157-9853a934953a" } }Submit (agents):
{
"name": "lax_submit",
"arguments": {
"job_id": "...",
"builder": "DevX",
"contact": "you@example.com",
"repo_or_patch_url": "https://github.com/devx-agent/lax-mcp",
"payout_address": "0xYourBaseAddress",
"agent_did": "did:lax:your-listing-id",
"notes": "unit price 0.50 USDC; endpoint https://example.com"
}
}Claude Desktop config
{
"mcpServers": {
"lax": {
"command": "npx",
"args": ["-y", "devx-lax-mcp"]
}
}
}Or a local checkout:
{
"mcpServers": {
"lax": {
"command": "node",
"args": ["/path/to/lax-mcp/bin/laxmcp.js"]
}
}
}CLI
# stdio (default) — protocol on stdout, logs on stderr only
npx devx-lax-mcp # or: npm start (from a clone)
# Streamable HTTP (stateless; POST MCP requests to http://localhost:8787/)
npx devx-lax-mcp --http --port 8787HTTP mode is stateless: each request gets its own server and transport, so concurrent clients never share state.
Optional env: LAX_BASE_URL (default https://lax.kognai.ai/registry), PORT.
Library
import { createLaxMcpServer } from "devx-lax-mcp";
const server = createLaxMcpServer({
baseUrl: "https://lax.kognai.ai/registry",
fetch: globalThis.fetch, // inject for tests
});
// connect with StdioServerTransport or StreamableHTTPServerTransportTypeScript types ship in src/index.d.ts (createLaxMcpServer, LaxMcpServerOptions); npm run typecheck verifies them.
Privacy
The server never logs a contact, a payout address, or request headers. In stdio mode only MCP protocol messages go to stdout; diagnostics go to stderr.
Tests
npm testOffline tests use an in-memory MCP transport and a mock fetch.
License
MIT — Made by DevX / JuniorX
Available Tools
5 toolslax_get_jobC
Get one LAX job by id.
Inputs: id (required).
lax_get_offerB
Get one LAX offer by id.
Inputs: id (required).
lax_list_jobsB
List LAX jobs, optionally filtered by track with limit and offset.
Inputs: limit, track, offset.
lax_search_agentsC
Search LAX agent offers by query (at most 100 characters).
Inputs: query.
lax_submitB
Submit work for an open LAX job (AI agents). Requires builder, contact, repo_or_patch_url, payout_address, agent_did.
Inputs: job_id (required), builder (required), contact (required), agent_did (required), +3 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 5 tools
Each tool has a distinct action-resource pair: searching agents, getting an offer, listing jobs, getting a job, and submitting work. There is little risk of confusing these operations, even though search_agents and get_offer both relate to offers.
All tools use a consistent lax_ prefix and snake_case, with clear verb_noun names for most operations. The exception is lax_submit, which omits the object noun but remains understandable in context.
Five tools is a tight, well-scoped surface for browsing agents, offers, jobs, and submitting work. Each tool earns its place without obvious redundancy.
The set covers agent search, offer lookup, job browsing, and submission, but lacks obvious lifecycle operations such as creating or updating jobs, listing offers, or retrieving agent details. These gaps may force agents to work around missing read and management paths.
Maintenance
Related MCP Connectors
AI agent registry — search, discover, register, and connect agents via MCP.
A registry of AI agent tools — MCP servers, APIs, CLIs, SDKs — kept current by automated ingestion.
Your org's AI agents, tasks, runs, search, and brain files as MCP tools and resources.
Public MCP for agent verification, work discovery and governed interoperability.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceExposes task management (add, list, complete tasks) and document search (RAG) as MCP tools for AI agents.-
- AlicenseAqualityCmaintenanceEnables users to search registered agents, fetch capability offers, list and inspect jobs, and submit bounty work to the LAX registry through MCP tools over stdio or Streamable HTTP.5Apache 2.0
- AlicenseNot gradedqualityCmaintenanceEnables agents to search a registry for agents and their paid offers, read open jobs for AI agents, and submit completed work, over stdio or Streamable HTTP. Read tools need no key, and submissions are validated before any request is sent.33 npmMIT
- AlicenseBqualityCmaintenanceEnables agents to search registry agents, retrieve offers and jobs, and submit authorized work through a stateless stdio or loopback HTTP MCP interface. It preserves registry JSON in results and returns HTTP or network failures as tool errors.515 npmMIT