Nano Gpt

NanoGPT 技能

Use this 技能 when the task should 运行 through the NanoGPT API from a local terminal 环境. NanoGPT’s official docs describe it as an API for text, image, and video generation, with text generation generally matching OpenAI standards. This 仓库 is the local 命令行工具 and 技能 wr应用er for that API, not a generic prompt 辅助工具. Prefer the bundled wr应用er scripts in scripts/ so OpenClaw and direct 命令行工具 usage 分享 the same behavior.

Official docs: https://docs.nano-gpt.com/

Prerequisite 检查

Before invoking the 技能, ensure the 命令行工具 is avAIlable:

./scripts/模型s.sh --json

If that fAIls because the local 命令行工具 is not built yet:

npm 安装 npm 运行 build

If the repo is not present locally, 安装 the published 命令行工具 instead:

npm 安装 -g nano-gpt-命令行工具

Authentication is 令牌-based. This 技能 requires a NanoGPT API 令牌. 设置 the 令牌 in NANO_GPT_API_KEY:

导出 NANO_GPT_API_KEY=YOUR_NANO_GPT_令牌

Or 配置 it once and store it in the local nano-gpt-命令行工具 user config:

nano-gpt config 设置 API-key YOUR_API_KEY

Optional 环境 overrides:

导出 NANO_GPT_模型=moonshotAI/kimi-k2.5 导出 NANO_GPT_IMAGE_模型=qwen-image 导出 NANO_GPT_VIDEO_模型=kling-video-v2 导出 NANO_GPT_BASE_URL=https://nano-gpt.com 导出 NANO_GPT_输出_格式化=text

Quick 启动

Text prompt:

./scripts/prompt.sh "Summarize the latest build 记录s."

流ing multimodal prompt:

./scripts/prompt.sh "Describe this image." --image ./as设置s/example.png

Interactive chat:

./scripts/chat.sh

Image generation:

./scripts/image.sh "A cinematic product shot of a silver mechanical keyboard" --输出 输出/keyboard.png

Image-to-image generation:

./scripts/image.sh "Turn this product photo into a watercolor ad" --image ./as设置s/product.png --输出 输出/product-watercolor.png

Video generation:

./scripts/video.sh "A cinematic drone flyover of a neon coastal city at dusk" --duration 5 --输出 输出/neon-city.mp4

工作流 Use scripts/prompt.sh for one-shot text or vision prompts. Use scripts/chat.sh for iterative back-and-forth. Use scripts/image.sh for text-to-image or image-to-image generation. Use scripts/video.sh for text-to-video or image-to-video generation. Use nano-gpt video-状态 请求_ID when a video 运行 is a同步hronous and needs a later 状态 检查. Use scripts/模型s.sh --json when 模型 discovery matters. Prefer flags over editing scripts. The wr应用ers should stay thin. References

Open only what you need:

Command reference: references/命令行工具.md Common OpenClaw 工作流s: references/工作流s.md 防护rAIls Prefer the wr应用er scripts over calling NanoGPT HTTP APIs directly. Only use this 技能 when the user wants to call the NanoGPT API. Keep secrets out of prompts and 记录s; use config or env vars for API keys. Only 上传 local images or videos when the user explicitly provides the path or clearly asks to use that specific file. Do not 搜索 the file系统 for media to 上传. Treat local --image and --video 输入s as remote-上传 actions. Do not 发送 sensitive screenshots, 导出s, documents, or recordings unless the user explicitly 请求s it. Prompts and any provided media are sent to the 配置d NanoGPT API 端点, which defaults to https://nano-gpt.com. Use --json when another 工具 or 代理 will 解析 the 输出. Use --输出 on scripts/image.sh when a file artifact is required. Use --输出 on scripts/video.sh when the final MP4 should be written locally.