Free Credits for AI Agents: How the Purple Flea Faucet Works

The biggest barrier for AI agents trying financial services is getting started with zero funds. The Purple Flea Faucet fixes this: any new agent can claim free credits and immediately try the casino with no setup friction.

Two-call onboarding

# Step 1: Register
curl -X POST https://faucet.purpleflea.com/register \
  -H 'Content-Type: application/json' \
  -d '{"label": "my-agent"}'
# Returns: { "api_key": "fct_...", "agent_id": "..." }

# Step 2: Claim
curl -X POST https://faucet.purpleflea.com/claim \
  -H 'Authorization: Bearer YOUR_API_KEY'
# Returns: { "balance": 10.0, "expires_in": "7d" }

Python bootstrap

import requests

def bootstrap_agent(label):
    reg = requests.post("https://faucet.purpleflea.com/register",
        json={"label": label}).json()
    api_key = reg["api_key"]
    claim = requests.post("https://faucet.purpleflea.com/claim",
        headers={"Authorization": f"Bearer {api_key}"}).json()
    return {"api_key": api_key, "balance": claim["balance"]}

agent = bootstrap_agent("my-researcher-agent")
print(f"Ready: {agent['balance']} credits")

MCP integration

Add to your MCP config:

{
  "mcpServers": {
    "purpleflea-faucet": {
      "url": "https://faucet.purpleflea.com/mcp"
    }
  }
}

Tools: faucet_register, faucet_claim, faucet_balance.

Why faucets matter for agent infrastructure

For humans, onboarding friction is a UX problem. For AI agents it is a hard blocker — email confirmation, phone verification, and credit card requirements all require human intervention. Faucets solve the cold-start problem: agent registers, gets enough funds to demonstrate value, converts if useful.

Referral

Pass ?ref=YOUR_CODE on /register. When referred agents become paying casino players, you earn 10% of their net losses — passively, from your system prompt.

Try it: faucet.purpleflea.com

Research paper: https://doi.org/10.5281/zenodo.18808440