AI BriefingAdvisor for Local Models
Key point
Open-Advisor supports the `/advisor` slash command across multiple AI CLIs.
Details
Open-Advisor is an open-source /advisor plugin that works inside Qwen CLI, Claude Code, Codex CLI, Gemini CLI, and OpenCode.
- Advisory targets include ChatGPT, Claude, Kimi, Qwen web versions, Claude Code, Codex CLI, Gemini CLI local CLIs, and OpenRouter-based models.
- During installation, you choose the host with
--ai <host>, and files are placed according to each environment's command directory and context file. - Installation for Qwen additionally handles MCP package and config file integration.
The usage flow is simple.
- Install with
node install.mjs --ai <host>. - Select an advisory target with
/advisor.select <name>. - Complete web account login or CLI binary verification with
/advisor.setup, then run/advisor.
Responses are saved to advisor-last-response.md in each host directory, with only a short preview shown in the body. Web advisory uses Chrome and Playwright profiles, while CLI advisory calls local binaries via spawnSync. OpenRouter requires an API key and receives the full response without asynchronous streaming.
Stated limitations are also noted. Web advisory has the overhead of launching Chrome, and CLI advisory is subject to a 90-second timeout.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Toss Doctor V2 Released: Maximizing QA Efficiency with LLM-based Automated Testing and Self-healing