GPT with Taste: System Instructions and Steering for Coding Harnesses

RepositoryNB-Corp/gpt-with-tasteTool

The open-source project gpt-with-taste provides tailored system instructions and event-based steering for AI coding agents. Designed primarily for the GPT reasoning family such as GPT-6.1 Sol and GPT-6 Astra, it aims to guide models toward better judgment and craftsmanship with less conversational ceremony.

Key points

  1. Compiles a single behavioral recipe into native instruction sets and event steering hooks across target hosts.
  2. Supports harnesses including OpenAI Codex, Claude Code, Kiki, Pi, Oh My Pi (OMP), and DeepSeek Harness (dsh).
  3. Requires Node.js 22 or newer for setup scripts without overriding MCP configurations, roles, or AGENTS.md.
  4. Leverages hooks in environments like Codex Desktop to inject scoped reminders at key execution steps.

Why it matters

AI coding assistants often generate verbose boilerplate or make arbitrary implementation choices. Standardising agent behavior and injection hooks across tools helps teams maintain consistent code craft and decision quality.

Analysis

gpt-with-taste addresses the gap between raw model reasoning and disciplined software engineering. Built on JavaScript and a compiler-based workflow, it maintains a single source recipe that adapts to various agent frameworks. The project is aimed at software engineers looking to refine automated agent workflows in tools like Codex and Claude Code. With roughly 350 stars and recent commits, it demonstrates a functional niche tool, though users should review its custom licensing status before organizational adoption.

Mentioned links

Found in

Summary generated by AI (Gemini 3.8 Flash, 18 October 2026). It may contain errors – the original source is authoritative.

Related entries

Similar in content, found by the AI search.

  • Repository:Leonxlnx/tastecode

    TasteCode: Local Desktop Workspace for AI Coding Agents

    TasteCode is an open-source desktop application for Windows and macOS that combines multiple AI coding agents such as Claude Code, Codex, and Grok into a unified interface. The tool follows a local-first security architecture and includes a dedicated Design Mode for editing web interfaces.

    367starsTypeScript

    AI & LLMs· Tool

  • Repository:GonChen/real-engineering-prompt

    Real Engineering Prompt: Minimal Behavioural Directives for Coding Agents

    This open-source project offers a concise instruction template (AGENTS.md) designed to bind modern coding agents to core engineering practices like minimal complexity, scoped modifications, and evidence-backed verification.

    4stars

    AI & LLMs· Resource

Back to the knowledge base

Let's talk about your project

Get in touch

Locations

  • Mattersburg
    Johann Nepomuk Bergerstraße 7/2/14
    7210 Mattersburg, Austria
  • Vienna
    Ungargasse 64-66/3/404
    1030 Wien, Austria

Parts of this content were created with the assistance of AI.