IvyeaOps

What IvyeaOps does

IvyeaOps is a self-hosted workspace for Amazon sellers, with a FastAPI backend and a React and Vite frontend served from the same server. One login gives access to a listing workbench, image translation, market research, deep analysis tools, advertising optimisation with Lingxing ERP integration, a knowledge base, a skill workshop and server operations. The README is written in Chinese.

A built-in agent called IvyeaAgent is installed by default and offers a chat entry point, knowledge-base upload and local search. External CLI agents such as Hermes, Claude and Codex can also be connected, while Hermes and Ollama are kept only as legacy compatibility components. Shop data and API keys stay in the backend data/ directory, which is gitignored, and controlled Lingxing write actions are read-only by default. It listens on port 8001 and is licensed AGPL-3.0.

Key features

• Listing workbench for titles, bullet points and descriptions, with image-set storyboards and per-image quality checks
• Market research, ASIN deep audits, ad search-term diagnosis and competitor tools
• Lingxing ERP integration through its official OpenAPI, with write actions guarded and read-only by default
• Built-in IvyeaAgent with a knowledge base, plus external CLI agents such as Hermes, Claude and Codex
• Skill workshop that generates executable skills from a one-sentence description

When to use it

• Running Amazon listing, research and advertising workflows from one self-hosted server
• Connecting an external CLI agent to an Amazon operations workspace
• Keeping shop data and API keys on your own server instead of a third-party cloud

Who it is for: Amazon sellers and operators who want a self-hosted AI workspace; the documentation is in Chinese.

How it fits with Hermes Agent

IvyeaOps can connect external CLI agents including Hermes, Claude and Codex, but keeps Hermes only as a legacy compatibility component; the built-in IvyeaAgent is the default.

Requirements: Linux or macOS (full support) or Windows (all but the PTY terminal); the Windows x64 and macOS release packages need no Python or Node.

Note: Hermes is kept only as a legacy compatibility component, and the documentation is written in Chinese.

FAQ