nativeprompt is a CLI tool that automatically rewrites user prompts into the native dialect of the target AI model's CLI (Claude Code or Codex) by applying that vendor's official, documented rules and showing exactly which rule drove each edit.

Stars

28

7-day growth

No data

Forks

2

Open issues

0

License

MIT

Last updated

2026-07-30

AI repository intelligence
FR-AI / ANALYSIS

Why it is worth attention

It fills a gap between competing vendor prompt recommendations, auto-detects the model you're running (without an API key), provides transparent, source-linked explanations for every change, and operates with zero runtime dependencies deterministically and offline.

Who it is for

  • Developers and engineers using Claude Code or Codex CLIs in their daily workflow
  • Prompt engineers who need to optimize prompts for different model families
  • Teams maintaining multi-model agentic CLIs and seeking consistent prompt engineering
  • Open-source contributors interested in rule-based prompt transformation tools

Use cases

  • Transforming a prompt tuned for Claude Code into one optimized for Codex (or vice versa) with documented reasoning
  • Ensuring team prompts comply with each vendor's latest published best practices
  • Integrating prompt improvement into CI/CD pipelines via the deterministic, offline interface
  • Auditing and debugging why a prompt performs differently on two model CLIs

Strengths

  • Zero runtime dependencies (stdlib only), works entirely offline and deterministically
  • Auto-detects the active model from the environment, session, or config files with transparent precedence
  • Every rule is sourced from official vendor documentation (linked) and explained per edit
  • Self-update mechanism (`update --diff`) shows exact changes in vendor docs, with weekly CI to detect drift

Considerations

  • Currently only supports Claude Code and Codex CLIs, not API or web chat interfaces
  • Detectors are regex heuristics, so they may produce false positives or miss subtle phrasing
  • CLI explanations are in Russian (English output on roadmap); alias-to-version resolution is limited

README quick start

Quick start

Description

Rewrite your prompt into the native dialect of the model you're actually running (Claude Code / Codex) — by each vendor's OFFICIAL rules — and see which rule drove each edit. Self-updating. Zero deps.

Related repositories

Similar projects matched by category, topics, and programming language.

lopopolo
Featured
lopopolo GitHub avatar

harness-engineering

Harness Engineering is a methodology for improving coding agent outputs by carefully crafting the environment around them—providing curated context, tools, and executable constraints that encode an organization’s nonfunctional requirements and cumulative lessons.

AI & Machine LearningAI Agents
2,390
MoonshotAI
Featured
MoonshotAI GitHub avatar

Kimi-K3

Kimi K3 is an open-weight, 2.8T-parameter native multimodal agentic model with a 1M-token context window, designed for frontier coding, knowledge work, and reasoning tasks.

AI & Machine LearningAI Agents
3,348
mshumer
Featured
mshumer GitHub avatar

Claude-of-Duty

A browser-based first-person shooter built entirely with procedural generation and orchestrated AI agents, featuring 55k lines of Three.js/WebGL2 code and no art assets.

AI & Machine LearningAI Agents
1,234