#199199th of 455 in Skills, plugins and prompts
langgpt
Structured prompt template framework with examples and a Claude Code skill
- Stars
- 12.7k
- License
- Apache-2.0
- Last commit
- Sep 2026
- Language
- Jupyter Notebook
11,776 on 21 Apr 202612,692 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
LangGPT is a Markdown template format for writing LLM prompts, with sections for Role, Profile, Goal, Skills, Rules, Workflow and Initialization, plus variables, slash commands and conditional logic. The repo holds guides, essays, example prompts and links to a Claude Code skill (structured-prompt-writer) and ChatGPT/Kimi prompt generators. It is documentation and prompt text rather than a runnable application.
Who it is for: Prompt writers and Claude Code users wanting reusable prompt templates
Strengths
- Template is plain Markdown and works with any chat model, no install
- Includes a Claude Code skill installable via /plugin marketplace add
- Backed by a paper (arXiv:2402.16929) and a tutorial set
- Example library includes FitnessGPT and a coding assistant prompt
Weaknesses
- Much of the essay and doc content is in Chinese
- The 100+ prompt library is hosted on Feishu, not in the repo
- Adoption and model-familiarity claims come with no benchmarks in the README
- README badge says MIT while detected license is Apache-2.0
What it needs
- no GPU
- Needs Claude Code (optional, for the skill)
- Models: ChatGPT, Claude, DeepSeek, Gemini, Kimi
Also in Skills, plugins and prompts
See all 455| Rank | Project | Score |
|---|---|---|
| 1 | claude-memPersistent cross-session memory plugin for Claude Code and other coding agents | 97 out of 100 |
| 2 | impeccableDesign skill with 24 slash commands for AI coding agents' frontend output | 97 out of 100 |
| 3 | gstackMarkdown slash-command skills that give Claude Code planning, review, QA and ship roles | 96 out of 100 |
| #4 | rufloAgent harness adding swarms, memory and hooks to Claude Code and Codex | 96 out of 100 |
| #5 | cavemanShortens coding-agent replies and compresses tool output before the model reads it | 95 out of 100 |