write-readme · git:20261115.db8ba56 · 2026-11-15 · sha256 17d8f1ce928ac266
write-readme git:20261115.db8ba56A
Immutable. This exact content never changes and is served at /api/v1/blob/17d8f1ce928ac266.
No tool to install first
claude plugin marketplace add https://markdownregistry.com/mp/a/art_nnqc4mog3n2z7js3/17d8f1ce928ac26644a7fda54386c3fef6e78c169fbb92f89392ab935b8d3608.json
Then /plugin install write-readme@markdownregistry-nnqc4mog3n2z7js3 in Claude Code. It installs this exact version, the one graded A here, pinned to the commit it was graded at, and leaves it there.
---
name: write-readme
description: Writes or updates a project's README.md to match a fixed template with badges, highlights, directory overview, system architecture and dev notes. Use when the user asks to write, rewrite, update or tidy a repository's README, or to document a project for GitHub. Examples include "write a README for this repo", "update my README to the usual template", "document this project" and "my README is out of date".
---
Prior to performing task, check that the github repository is attached to this chat, otherwise, prompt the user to attach repository.
Update the attached github repository's README.md file. Your README.md file must align with the following template example below.
# README.md Template
````markdown
# Project Title
[ensure to include shields.io badges alike examples below]
[](LICENSE)
[](https://python.org)
A concise, **one-to-two sentence description** of what this project does, the problem it solves relative to what my project's main objective and methodology.
## Highlights
- **Objective**: [main objective of the entire project]
- **Key Feature** [input the feature/method/intent/agenda of this project. If multiple features then please list in summary points]
- **Tech stack** [input how the key feature/s was/were implemented in this project e.g. depended open source library, dataset used if any,.etc]
* **Evaluation** [input how the idea/feature/model/algorithm was tested, any GPU acceleration, parallelization, or speed benchmarks]
* **Results & Conclusion** [list maximum of 2-3 summary points of the conclusion and decision made based on the idea implementation. Also, list extension to this project in moving forward]
## Project Directory Overview
[use CLI `tree` e.g. `tree -L 4 -d -I '__pycache__|.venv|node_modules' >` to return the subdirectory structure in the following example format]
Example project structure
```text
├── data/ # Raw and processed scientific data
├── src/ # Core source code and algorithms
├── tests/ # Unit tests
├── config.yaml # Hyperparameters and file paths
└── main.py # Main execution script
```
````
## System Architecture
## Dev Notes
- **Requirements**
- [list the requirements if its currently mentioned in existing README, otherwise, you may skip this part]
- **Installation**: [input bash code of CLI steps installing the Git repo alike example below]
```bash
# Clone the repository
git clone https://github.com
cd project-name
# Create and activate environment
conda create -n sci-env python=3.10 -y
conda activate sci-env
# Install dependencies
pip install -r requirements.txt
```
- **To start**: [input bash code of CLI steps starting/executing the project]
```bash
python main.py --config config/demo.yaml
```
- **Test**: [input bash code of CLI steps to run checks and tests on project]
- [You may skip this part if there was no research paper mentioned throughout the source code. Otherwise, refer to examples below]
- **Reproducing Results**: To reproduce the primary findings or figures stated in the paper/documentation:
1. Update paths in `config.yaml`.
2. Run the training/analysis pipeline:
```bash
python src/pipeline.py --mode train
```
## Citation
If you use this software or method in your research, please cite it as follows:
```bibtex
@article{author2026title,
author = {Lastname, Firstname and Collaborator, Second},
title = {Scientific Project Title},
journal = {Journal Name},
year = {2026},
doi = {10.1000/xyz123}
}
```