ProxyLLM Account Setup for Openclaw

Openclaw Skills for ProxyLLM lets an agent create, verify, and configure a gateway account, then provision fallback routing keys without a browser.

ericslab
v1.0.0
Jul 15, 2026
0
329
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install proxyllm-account

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install proxyllm-account using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is ProxyLLM Account Setup?

Openclaw Skills for ProxyLLM turns LLM gateway onboarding into an autonomous, API-first workflow: an agent signs up with an operator email, verifies the OTP, handles the expected paywall handoff, and provisions OpenAI-compatible routing keys.

The result is a managed ProxyLLM account with fallback lanes, budgets, and request logs, so your agent can keep using standard OpenAI SDKs while routing traffic through provider lanes such as a flat ChatGPT/Codex subscription, a self-hosted bridge, or metered API keys.

ProxyLLM Account Setup Use Cases

  • Bootstrapping LLM access for an AI agent without using a browser or solving captchas.
  • Converting an operator's paid ChatGPT/Codex subscription into a routed fallback lane for API-style usage.
  • Creating metered, revocable keys for sub-agents with per-key monthly budgets.
  • Adding provider fallback across Codex, self-hosted bridges, and saved API keys.
  • Auditing LLM requests with live logs, model discovery, and usage reporting.
  • Exposing an OpenAI-compatible base URL to existing SDKs and tools.
  • Connecting the same surface through MCP for agent-native integrations.

How ProxyLLM Account Setup Works

  1. Submit an agent signup request with the operator's email to the ProxyLLM auth endpoint.
  2. Read the 6-digit OTP from the inbox, then verify it to receive the one-time sk_... account token.
  3. If the API returns 402 payment_required, relay the checkout message and payment link to the operator, then poll organization status until the plan becomes active.
  4. Create a routing key, inspect available lanes, and PATCH the key to bind one or more providers and models in fallback mode.
  5. Optionally connect a Codex subscription on the operator machine, then export OPENAI_BASE_URL and the routing key as OPENAI_API_KEY.
  6. Use the key with any OpenAI-compatible client, or connect through the MCP server for agent workflows.

ProxyLLM Account Setup Setup

Prerequisites

  • An operator email that can receive OTP messages.
  • A secure place to store secrets such as SK_TOKEN and OPENAI_API_KEY.
  • Optional access to the operator's machine if a Codex lane will be connected.

1) Create the account

curl -X POST "https://proxyllm.ai/api/auth?action=agent-signup" -H "Content-Type: application/json" -d '{"email":"OPERATOR_EMAIL","source":"clawhub"}'

2) Verify the OTP and capture the management token

curl -X POST "https://proxyllm.ai/api/auth?action=agent-verify" -H "Content-Type: application/json" -d '{"email":"OPERATOR_EMAIL","code":"123456"}'

Store the returned account_token (sk_...) exactly once in your secret manager.

3) Handle the expected paywall handoff

If the response is 402 payment_required, send the operator the checkout message and checkout URL, then poll until the plan flips from free to pro.

curl -s https://api.proxyllm.ai/v1/organizations/me -H "Authorization: Bearer $SK_TOKEN"

4) Mint a routing key and export client settings

curl -s -X POST https://api.proxyllm.ai/v1/organizations/routing-keys -H "Authorization: Bearer $SK_TOKEN" -H "Content-Type: application/json" -d '{"label":"agent"}'

export OPENAI_BASE_URL=https://api.proxyllm.ai/v1
export OPENAI_API_KEY=pllm_...

5) Optional: install the MCP server

npx -y proxyllm-mcp

Use the same OpenAI-compatible credentials with any existing SDK, and add provider lanes or budgets as needed.

ProxyLLM Account Setup Data Schema & Taxonomy

Core entities and identifiers

Artifact Prefix / Endpoint Purpose Important fields
Account token sk_... Full management token returned once at verify time account_token, revocable, secret
Routing key pllm_... OpenAI-compatible client credential label, mode, monthly_budget_usd, provider bindings
Provider lane codex, bridge, or saved key Route target for fallback fan-out provider, model, codex_session_id
Organization /v1/organizations/me Membership and entitlement state plan, account_expires_at, usage
Usage log /v1/organizations/usage 30-day usage reporting request totals, model usage, budget consumption
Model catalog /v1/models Live model discovery no-auth catalog
Payment payload 402 payment_required body Human relay instructions payment.checkout_url, payment.price_usd, message

Local config and generated secrets

  • No mandatory local files are created by the skill.
  • Commonly stored values:
    • SK_TOKEN for organization management
    • OPENAI_BASE_URL=https://api.proxyllm.ai/v1
    • OPENAI_API_KEY=pllm_...
  • If you prefer local persistence, keep these in a .env file or a secret manager entry.
  • Optional operator notes may include the checkout URL and payment status while the account is pending.

ProxyLLM Account Setup Advanced Features

  • OpenAI-compatible drop-in endpoint with unchanged SDK usage.
  • Fallback routing across multiple provider lanes and models.
  • Budget controls at the key level via monthly_budget_usd.
  • Full request logs and usage introspection for auditing.
  • Human-in-the-loop payment handling instead of browser automation or captcha solving.
  • Optional Codex bridge support using codex login and proxyllm codex connect.
  • MCP integration via npx -y proxyllm-mcp.
  • Scoped routing keys that can be delegated to sub-agents.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*