Turn a rough, informally described requirement into a precise, executable prompt for Codex
Turn a Chinese-language requirement into a structured execution prompt for Claude Code
Install the skill as a Claude Code plugin to invoke it with a slash command
Use the plugin marketplace to distribute prompt-mode alongside other Codex plugins
| jasonz3g/prompt-mode | 0xkinno/neuralvault | 0xlocker/d17-contracts | |
|---|---|---|---|
| Stars | 1 | 1 | 1 |
| Language | — | TypeScript | Solidity |
| Setup difficulty | easy | hard | hard |
| Complexity | 1/5 | 4/5 | 5/5 |
| Audience | developer | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Install via Codex's skill installer or Claude Code's plugin-dir flag, no external dependencies mentioned.
Prompt Mode is a skill for AI coding assistants that turns a rough, informally described requirement into a clear, ready-to-use prompt for another AI agent to carry out. Rather than doing the underlying task itself, it focuses only on producing the instructions that describe how the task should be done. When used, it analyzes what the user actually wants, looks at relevant local project files and documents the user mentions, and picks out which of the assistant's available skills would actually be useful for that task. It also adds stop points, boundaries, and expectations for how the result should be verified, so the resulting prompt is more precise than a quick, informal request would be. Notably, the README states that it deliberately refuses to silently go ahead and execute the task itself while it is supposed to be turning that task into a prompt, keeping the two steps separate. The repository is structured as a plugin that can be installed in two different AI coding tools. In Codex, it can be installed as a standalone skill using Codex's built in skill installer, or the whole repository can be added as a plugin marketplace so the prompt-mode plugin can be installed from there. In Claude Code, it can be tested locally by pointing Claude Code at the plugin directory after cloning the repository, or installed for personal use by copying or linking the skill folder into the user's Claude Code skills directory, after which it can be invoked with a slash command. Once installed, it is triggered with specific phrases, including examples in both English and Chinese, such as asking it to rewrite a requirement into an executable prompt for Codex, or asking it in Chinese to organize a rough requirement into an execution prompt for Claude Code. The project is released under the MIT license.
A skill for AI coding assistants that turns a rough, informal task description into a clear, ready-to-use prompt for another agent to execute.
Licensed under MIT, so anyone can freely use, copy, modify, and share it, including for commercial purposes, as long as the copyright notice is kept.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.