add-newcli-provider

What This Skill Does

This skill, add-newcli-provider, configures OpenClaw to use code.newcli.com as a model source. It sets up four distinct providers: newcli for the main Claude line, newcli-aws for a special low-cost Claude line (consuming 1/24th the tokens), newcli-codex for GPT series models, and newcli-gemini for Gemini series models. This skill handles the complete process of provider registration, model definition, alias configuration, and fallback linking and verification. It is designed to be triggered by administrator commands such as "add Claude," "add GPT," "configure newcli," "add fox source," "integrate Claude model," "integrate GPT model," "add codex," or "add aws line."

NewCLI acts as a model proxy service, providing access to various LLMs through a unified account and API key. The four providers are necessary due to different API protocols and endpoints used by Claude, GPT, and Gemini. Additionally, Claude has two distinct lines (newcli and newcli-aws) with different token consumption rates, requiring separate configurations. The newcli-aws provider is particularly noteworthy for its significantly reduced token cost, making it ideal for important tasks, with the main newcli provider serving as a fallback.

Installation

To install this skill, use the following command:

clawhub install openclaw/skills/skills/jooey/add-newcli-provider

Use Cases

• Integrating diverse LLM backends: Easily add support for Anthropic Claude, OpenAI GPT, and Google Gemini models through a single proxy service.
• Cost optimization: Leverage the newcli-aws provider for significant cost savings on Claude model usage.
• Simplified model management: Configure multiple model providers with a single skill, streamlining the OpenClaw setup process.
• Flexible AI backend selection: Enable users to choose between different model providers and lines based on cost, performance, or availability.
• Automated provider setup: Automate the often tedious process of registering new model providers, including defining models, aliases, and fallback mechanisms.

Example Prompts

1. "Configure newcli with my API key and Claude, GPT, and Gemini models."
2. "Add the AWS Claude line as the primary source for newcli."
3. "Set up codex and gemini providers using code.newcli.com."

Tips & Limitations

• API Key: Ensure you have a valid API key from NewCLI. This key is shared across newcli, newcli-aws, and newcli-codex providers.
• Model Verification: It is crucial to verify which models are actually available and working with your API key. Do not rely solely on documentation; test models using the provided curl commands or similar tools.
• AWS Line Usage: The newcli-aws line offers a 1/24 token consumption rate. It is recommended to prioritize this line for important tasks.
• Cloudflare Protection: The /v1/models endpoint on NewCLI is protected by Cloudflare, preventing direct API calls to list available models. Always confirm model availability through your NewCLI account dashboard.
• Provider Differences: While newcli and newcli-aws use the same API key and serve Claude models, they have different base URLs and, importantly, different cost structures. The newcli-codex and newcli-gemini providers use distinct API protocols and endpoints.
• Security Warning: The provider explicitly warns against distributing the AWS line address, as doing so may result in account suspension and forfeiture of funds.