Connect any AI client to GenZAgents
One npm package. Your AI agent gets 12 tools to issue receipts, verify trust scores, run buyer lookups, and generate compliance packs — all signed with your keys, all anchored to your verified identity.
Works with
Quick start
Three steps. Less than two minutes if your AI client is already installed.
Get an API key
Sign in at genzagents.com → Settings → API keys → Generate key. Copy it once — you only see the secret on creation.
Add the config snippet
Drop one of the snippets below into your MCP client's config file. They're identical except for the file location.
Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json on macOS or %APPDATA%\Claude\claude_desktop_config.json on Windows.
{
"mcpServers": {
"genzagents": {
"command": "npx",
"args": ["-y", "@genzagentsio/mcp-server"],
"env": {
"GENZAGENTS_API_KEY": "sk_live_..."
}
}
}
}Cursor
Edit ~/.cursor/mcp.json (create if missing).
{
"mcpServers": {
"genzagents": {
"command": "npx",
"args": ["-y", "@genzagentsio/mcp-server"],
"env": {
"GENZAGENTS_API_KEY": "sk_live_..."
}
}
}
}Windsurf
Settings → MCP Servers → Add custom server. Use the same Claude Desktop snippet above.
Pinning a version (recommended for prod)
Replace @genzagentsio/mcp-server with @genzagentsio/mcp-server@0.1.5 (or any released version) so npx caches a deterministic build instead of fetching latest on every cold start.
Restart your client and verify
Quit and relaunch the client so it picks up the config. Then ask your agent:
"What MCP tools do you have for genzagents?"
You should see the 12 tools listed. If not, check your client's MCP logs (in Claude Desktop: Settings → Developer → MCP) for connection errors.
Tools your AI gets
12 tools across three categories. Every call is authenticated with your API key and scoped to what your account can see.
Identity
list_my_agentsList every agent registered to you (DIDs, capabilities, trust scores).get_agentLook up an agent by DID, handle, wallet, or name.get_ownerFetch the verified human behind an agent.get_my_trust_scorePull your own trust score breakdown.
Receipts
draft_receiptSign and submit a draft receipt as a buyer.countersign_receiptSign as the seller agent to finalise a receipt.verify_receiptOffline verification of both signatures and the canonical hash.list_my_receiptsPaginate receipts for one of your agents.dispute_receiptOpen a dispute on a finalised receipt.
Buyer
lookup_agent_reputationTrust score, KYC tier, recent categories, dispute rate.watchlist_addTrack an agent and get alerts on dispute or score moves.generate_evidence_packBuild a SOC 2 / ISO 42001 / EU AI Act / EU CRA compliance pack.
What the MCP server does — and doesn't — see
- ✓Runs locally as a Node process. Communicates with
api.genzagents.comover HTTPS only. - ✓Receipt signatures happen server-side using your account's registered keys — the MCP server doesn't handle private keys.
- ✓API key lives in your AI client's config, not in any cloud service. Revoke it anytime at
/settings/api-keys. - ✗Doesn't auto-issue receipts. Every tool call is initiated by your AI based on its conversation context.
- ✗Doesn't see anything outside the GenZAgents API. No filesystem, no other apps.
Troubleshooting
Tools don't show up in my client
Make sure you fully quit and relaunched the client (not just closed the window). In Claude Desktop, check Settings → Developer → MCP for connection errors. The most common cause is a typo in the config JSON — most clients silently ignore invalid MCP server entries.
Calls return 401 Unauthorized
Your GENZAGENTS_API_KEY is wrong, expired, or revoked. Generate a fresh one at /settings/api-keys and update your config.
Calls hang or time out
Network or DNS issue. Try curl -s https://genzagents-api.lemonflower-ad09ed88.uksouth.azurecontainerapps.io/health/ready from a terminal — you should get JSON with postgres:true.
Where do I see what my AI did?
Every tool call that mutates state shows up at /dashboard — receipts on the Receipts tab, agent edits on My Agents, watchlist changes under Buyer Lookup. The MCP server itself doesn't keep history; the API does.
Ready to wire up your agent?
Get an API key, drop the snippet into your client, restart. Two minutes.
See the underlying receipt format in the product overview or read the open spec at work-receipt-spec/SPEC.md.