Installable Agent Skills for Interfaze AI. This repo follows the open Agent Skills format and is compatible with the skills CLI.
Interfaze AI is a proprietary multimodal AI model accessible through an OpenAI-compatible API. It handles vision OCR, speech-to-text, structured output, object detection, web scraping, and document processing through a single unified endpoint — no tool routing or multi-provider setup required.
| Skill | Purpose |
|---|---|
| ocr | OCR and visual text extraction from images, scans, PDFs, invoices, receipts, and forms |
| speech-to-text | Speech-to-text transcription from audio files, voice notes, and recordings |
| structured-output | Convert model responses into strict JSON or schema-constrained structured objects |
| object-detection | Detect and locate objects in images with bounding box coordinates |
| web-search | Search the web for current information, facts, prices, and news |
| web-scraping | Extract structured data from specific web pages and URLs |
Install all skills:
npx skills add JigsawStack/interfaze-skillsInstall a single skill:
npx skills add JigsawStack/interfaze-skills --skill ocrList available skills without installing:
npx skills add JigsawStack/interfaze-skills --listInterfaze AI uses an OpenAI-compatible API. Set your API key and base URL:
import { createOpenAI } from "@ai-sdk/openai";
const interfaze = createOpenAI({
baseURL: "https://api.interfaze.ai/v1",
apiKey: process.env.INTERFAZE_API_KEY,
});Use the model name interfaze-beta with Vercel AI SDK functions like generateObject and generateText.
- AGENTS.md contains repo-wide rules and conventions that apply to all skills. Agents load this as passive context.
- SKILL.md files contain specialized instructions for a single capability. Agents load these on-demand when a task matches the skill description.
This repo follows the Agent Skills open specification. Skills work with 18+ AI agents including Claude Code, GitHub Copilot, Cursor, Cline, and others.
MIT — see LICENSE.