Floe Payments and Credit
Credit and payments for AI agent developers — over MCP. No crypto required
README
@floelabs/mcp-server
The Financial OS for AI Agents — exposed over MCP.
Connect Claude Desktop, Claude Code, Cursor, CrewAI, or any MCP-compatible client to the full Floe stack: wallet, on-ramp, working capital, x402 payments, and credit thresholds. 36 tools across the six components, with transport-aware auth (remote HTTP uses a Bearer token; local stdio reads FLOE_API_KEY from the env).
Proof points: 3,000+ secured working capital lines issued · zero defaults · 13,000+ x402 APIs reachable via the Floe proxy.
The Floe Stack (what this MCP server exposes)
| # | Component | Status | Tools |
|---|---|---|---|
| 01 | Agent Wallet | GA |
get_wallet_balance, get_credit_remaining, get_loan_state |
| 02 | Fiat on/off-ramp | Dashboard-driven | On-ramp links generated server-side; no MCP tool required today. Tool surface Roadmap. |
| 03 | Secured working capital | GA |
get_markets, get_market_details, get_open_lend_intents, get_open_borrow_intents, get_intent_details, get_loan, get_user_loans, get_loan_health, get_liquidation_quote, create_lend_intent, create_borrow_intent, create_counter_intent, repay_loan, add_collateral, withdraw_collateral, liquidate_loan, revoke_intent, approve_token, get_accrued_interest, get_token_price, check_compatibility, calculate_risk, estimate_interest |
| 04 | Unsecured working capital | Preview |
Coming soon — email hello@floelabs.xyz for the design partner program |
| 05 | x402 payment facilitator | GA (preflight + gating) |
estimate_x402_cost. Payment execution flows through https://x402.floelabs.xyz/proxy/fetch. |
| 06 | Credit & trust bureau | Writer GA · Portable reader Preview |
list_credit_thresholds, register_credit_threshold, delete_credit_threshold. Portable ERC-8004 reader tool is on the roadmap (see below). |
Plus utility tools — simulate_transaction, broadcast_transaction, get_transaction_status — shared across components.
Tested clients
| Client | Status |
|---|---|
| Claude Desktop | GA |
| Claude Code | GA |
| Cursor | GA |
| Continue / Cline | Best-effort |
CrewAI (via langchain-mcp-adapters) |
Beta |
| OpenAI Agents SDK | Preview (MCP fallback while native adapter ships) |
| ElizaOS | Preview |
Quick Start
Option 1: Remote (recommended)
Point your MCP client directly at the hosted endpoint — no installation needed.
Claude Desktop / Claude Code:
{
"mcpServers": {
"floe": {
"url": "https://mcp.floelabs.xyz/mcp",
"headers": {
"Authorization": "Bearer floe_YOUR_AGENT_KEY"
}
}
}
}
Option 2: Local via npx
Run the server locally. It proxies all requests to the Floe API.
FLOE_API_KEY=floe_YOUR_AGENT_KEY npx @floelabs/mcp-server
Claude Desktop config:
{
"mcpServers": {
"floe": {
"command": "npx",
"args": ["@floelabs/mcp-server"],
"env": {
"FLOE_API_KEY": "floe_YOUR_AGENT_KEY"
}
}
}
}
Cursor config (.cursor/mcp.json):
{
"mcpServers": {
"floe": {
"command": "npx",
"args": ["@floelabs/mcp-server"],
"env": {
"FLOE_API_KEY": "floe_YOUR_AGENT_KEY"
}
}
}
}
Option 3: Install globally
npm install -g @floelabs/mcp-server
FLOE_API_KEY=floe_YOUR_AGENT_KEY floe-mcp
Auth model
Auth source depends on transport:
| Transport | Identity source |
|---|---|
Remote HTTP (https://mcp.floelabs.xyz/mcp) |
Authorization: Bearer <key> header (per-request) |
Local stdio (floe-mcp / npx @floelabs/mcp-server) |
FLOE_API_KEY env var |
| Local HTTP (self-hosted) | Bearer header takes precedence; when ALLOW_SHARED_KEY_FALLBACK=true, the server falls back to FLOE_API_KEY if no header is sent |
Which key to use
Two key formats unlock different surfaces:
| Key format | Scope | When to use |
|---|---|---|
floe_<64-hex> (agent key, recommended) |
One specific agent | Default for MCP. Required for agent-awareness tools (get_credit_remaining, get_loan_state, get_spend_limit, etc). One MCP session = one agent. |
floe_live_<base62> (developer key) |
Whole developer account | Use only if you're running a multi-tenant integration that needs to see all agents. Agent-awareness tools return 401 because the caller is the developer, not a single agent. |
Get an agent key:
- Go to dev-dashboard.floelabs.xyz
- Connect your wallet and Create an agent (name + borrow limit + max rate)
- Copy the
floe_<64-hex>key shown at the end of the wizard — it is revealed once
You can also mint one from the CLI without visiting the dashboard:
# TypeScript SDK
npx floe-agent register --name my-agent --borrow-limit 10000
# Python SDK
floe-agent register --name my-agent --borrow-limit 10000
Get a developer key (only if you need multi-tenant access across all your agents):
- Go to dev-dashboard.floelabs.xyz/keys
- Click Create Key, label it, pick
readorread_writepermissions - Copy the
floe_live_<base62>key shown once
Developer keys span the whole developer account and have a separate rate limit (100 req/min). Agent-awareness tools (get_credit_remaining, get_spend_limit, etc) return 401 with a developer key because the caller is the developer, not a single agent — use an agent key for those. See the API Keys docs for the full taxonomy.
Fund with fiat: You can fund your wallet with USDC via Coinbase — credit card, bank transfer, Apple Pay, Google Pay — directly from the dashboard. No crypto on-ramp needed.
Multiple agents
One Floe developer can own many agents. To run several MCP sessions side by side (e.g. a research agent and a trading agent), mint one key per agent and configure each MCP client entry with its own key:
{
"mcpServers": {
"floe-research": {
"url": "https://mcp.floelabs.xyz/mcp",
"headers": { "Authorization": "Bearer floe_KEY_FOR_RESEARCH_AGENT" }
},
"floe-trading": {
"url": "https://mcp.floelabs.xyz/mcp",
"headers": { "Authorization": "Bearer floe_KEY_FOR_TRADING_AGENT" }
}
}
}
Each session is scoped to one agent — credit lines, spend limits, and webhooks stay isolated.
Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
FLOE_API_KEY |
Yes | — | Your Floe API key (floe_<64-hex> agent key recommended; floe_live_<base62> developer key also accepted) |
FLOE_API_BASE_URL |
No | https://credit-api.floelabs.xyz |
API endpoint |
MCP_PORT |
No | 3100 |
HTTP server port (non-stdio mode) |
ALLOW_SHARED_KEY_FALLBACK |
No | false |
Allow env-var key fallback when no Bearer header is sent (HTTP mode only) |
What can agents do?
| Floe component | Capability |
|---|---|
| Agent Wallet | Read balances, credit headroom, loan-lifecycle state |
| Secured working capital | Browse markets, post intents, match offers, repay, manage collateral, liquidate |
| x402 payment facilitator | Preflight x402 costs against current credit and spend limits |
| Credit & trust bureau | Register webhook thresholds on credit utilization; list / delete |
Tools (36)
Below the tools are listed by request type. Mapping to the six product components is in the table above.
Read tools
| Tool | Description |
|---|---|
get_markets |
List active lending markets with rates and liquidity |
get_market_details |
Detailed market info including oracle prices |
get_open_lend_intents |
Browse lend offers available for borrowing against |
get_open_borrow_intents |
Browse borrow requests from borrowers seeking lenders |
get_intent_details |
Get full details of a specific intent by hash |
get_loan |
Get loan details by numeric ID |
get_user_loans |
Get all loans for a wallet (borrower + lender) |
get_loan_health |
Check loan LTV, health status, liquidation risk |
get_liquidation_quote |
Get liquidation eligibility and details |
get_token_price |
Current oracle price for collateral tokens |
get_wallet_balance |
Token balances for a wallet |
get_accrued_interest |
Interest accrued on a loan |
Write tools (return unsigned transactions)
| Tool | Description |
|---|---|
create_lend_intent |
Create a lending offer |
create_borrow_intent |
Create a borrowing request |
create_counter_intent |
Accept an existing offer (solver matches automatically) |
repay_loan |
Repay a loan with slippage protection |
add_collateral |
Add collateral to improve loan health |
withdraw_collateral |
Withdraw excess collateral |
liquidate_loan |
Liquidate an unhealthy loan |
revoke_intent |
Cancel an active intent |
approve_token |
Approve token spending for the protocol |
Analysis tools
| Tool | Description |
|---|---|
check_compatibility |
Check if two intents can match |
calculate_risk |
Risk metrics: LTV, liquidation price, buffer |
estimate_interest |
Interest estimate for given loan terms |
Utility tools
| Tool | Description |
|---|---|
simulate_transaction |
Dry-run a transaction (eth_call) |
broadcast_transaction |
Submit a signed transaction |
get_transaction_status |
Check transaction receipt |
Agent-awareness tools
Lets an agent answer "do I have credit?", "is this call worth it?", and "where am I in the loan lifecycle?" before committing capital. All require an agent API key (floe_*). The calling identity is taken from the Bearer header in HTTP mode, or from FLOE_API_KEY in stdio mode (and as a fallback in HTTP mode when ALLOW_SHARED_KEY_FALLBACK=true).
| Tool | Description |
|---|---|
get_credit_remaining |
Current available credit, headroom to auto-borrow, utilization in bps |
get_loan_state |
Coarse state: idle | borrowing | at_limit | repaying |
get_spend_limit |
Currently active session spend cap, if any |
set_spend_limit |
Set a session-level USDC ceiling (resets the session window) |
clear_spend_limit |
Remove the session spend cap |
list_credit_thresholds |
List registered credit-utilization thresholds |
register_credit_threshold |
Register a webhook trigger at a utilization threshold (cap: 20 per agent) |
delete_credit_threshold |
Remove a registered threshold |
estimate_x402_cost |
Preflight an x402 URL — returns cost + reflection against your credit, no payment |
Roadmap tools (not yet shipped)
get_credit_profile— read a portable ERC-8004 credit record (Preview)request_unsecured_credit— apply for receivables-backed credit (Preview)create_onramp_link— generate a one-shot fiat on-ramp URL for an agent operator (Roadmap)
Email hello@floelabs.xyz for early access to any of these.
Transaction Flow
All write tools return unsigned transactions — the server never holds private keys.
1. Call a write tool (e.g., create_counter_intent)
→ Returns { transactions: [...], summary, warnings, expiresAt }
2. (Optional) Call simulate_transaction to dry-run
3. Sign each transaction locally with your wallet
4. Call broadcast_transaction with the signed hex
→ Returns { transactionHash, status, blockNumber }
Example: Get a USDC Credit Line
Agent: "I need 9,950 USDC working capital"
1. get_open_lend_intents → browse USDC/USDC offers
2. create_counter_intent(offer_hash, wallet) → unsigned txs
3. simulate_transaction(from, to, data) → { success: true, gasEstimate }
4. Sign locally → signed hex
5. broadcast_transaction(signed_hex) → confirmed
Signing with viem
import { createWalletClient, http } from "viem";
import { privateKeyToAccount } from "viem/accounts";
import { base } from "viem/chains";
const wallet = createWalletClient({
account: privateKeyToAccount(PRIVATE_KEY),
chain: base,
transport: http(),
});
// Sign and send each transaction in order
for (const { transaction: tx } of response.transactions) {
const hash = await wallet.sendTransaction({
to: tx.to,
data: tx.data,
value: BigInt(tx.value),
});
// Wait for confirmation before next step
}
Programmatic Usage
MCP Client SDK
import { Client } from "@modelcontextprotocol/sdk/client";
import { StreamableHTTPClientTransport } from "@modelcontextprotocol/sdk/client/streamableHttp.js";
const client = new Client({ name: "my-agent" });
await client.connect(new StreamableHTTPClientTransport(
new URL("https://mcp.floelabs.xyz/mcp"),
{ requestInit: { headers: { "Authorization": "Bearer floe_..." } } }
));
const markets = await client.callTool("get_markets", {});
const counter = await client.callTool("create_counter_intent", {
offer_hash: "0x...",
wallet_address: "0x...",
});
LangChain / LangGraph
from langchain_mcp_adapters import MultiServerMCPClient
async with MultiServerMCPClient({
"floe": {"url": "https://mcp.floelabs.xyz/mcp", "headers": {"Authorization": "Bearer floe_..."}}
}) as client:
tools = client.get_tools()
# Use tools in your agent
CrewAI
CrewAI agents can consume the Floe MCP tools via langchain-mcp-adapters. A runnable crew is available in floe-examples/crewai-demo.
Architecture
Your Agent → MCP Server → credit-api.floelabs.xyz → Envio Indexer / Base RPC
↑ ↑
This package Private backend
(open source) (holds secrets)
The MCP server is a thin HTTP client. All protocol logic, indexer queries, and RPC calls happen in the private Floe API backend. This package contains only tool definitions and fetch() calls.
Protocol overview
Floe is an intent-based lending protocol on Base, surfaced as the lending layer of the Financial OS:
- Primary market (USDC/USDC): Deposit USDC as collateral, borrow up to 99.5% as a credit line. No price-volatility risk — same-token market.
- Volatile markets: Also supports WETH and cbBTC collateral for crypto-native use cases.
- Solvers automatically match compatible intent pairs on-chain.
- Loans are created with matched terms, collateral locked in per-loan isolated escrow.
- Gas-free — Floe sponsors all transaction costs.
- Fixed rates — no variable-rate surprises.
Key concepts:
- Intent: An on-chain offer to lend or borrow
- Counter-Intent: An intent created to match an existing offer
- Health Factor: Ratio of collateral value to debt — below threshold triggers liquidation
- LTV (Loan-to-Value): Borrower's debt as % of collateral value
Contract Addresses (Base Mainnet)
| Contract | Address |
|---|---|
| LendingIntentMatcher | 0x17946cD3e180f82e632805e5549EC913330Bb175 |
| PriceOracle | 0xEA058a06b54dce078567f9aa4dBBE82a100210Cc |
| LendingViews | 0x9101027166bE205105a9E0c68d6F14f21f6c5003 |
| x402 Facilitator | 0x58EDdE022FFDAD3Fb0Fb0E7D51eb05AaF66a31f1 |
Links
- Website
- Dashboard
- Documentation
- TypeScript SDK (
floe-agent) - Python SDK (
floe-agentkit-actions) - End-to-end examples
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。