Skip to main content

A collection of prompts

Project description

bartste-prompts

A command-line tool to generate prompts for Large Language Models (LLMs) using customizable instruction templates.

Introduction

bartste-prompts is a CLI tool designed for developers to generate prompts for Large Language Models (LLMs). It uses a set of predefined instructions that can be customized per command and per file type. The tool is built with extensibility in mind, allowing users to define their own instructions. The CLI will adapt to custom instructions as a fixed directory structure is assumed.

Key features include:

  • Dynamic prompt assembly from markdown instruction templates
  • Support for multiple default commands (e.g., explain, fix, refactor, etc.) and file types
  • Integration with tools like aider for code editing and explanation
  • Can be extended with custom instructions
  • The CLI adapts to new instructions. No changes in code are necessary.
  • Question answering through the ask command for quick guidance

Installation

You can install bartste-prompts using pip:

pip install bartste-prompts

For development dependencies (testing, linting, etc.), install with:

pip install bartste-prompts[dev]

This will install the prompts command-line tool.

Usage

The basic command structure is:

prompts <command> [options]

Available Commands

The tool comes with these default commands:

  • ask: Answer user questions directly
  • docstrings: Generate docstrings for given files
  • explain: Explain code functionality
  • fix: Fix code issues
  • refactor: Improve code structure
  • typehints: Add type hints to code
  • unittests: Generate unit tests
  • commit: Generate conventional commit messages

Common Options

All commands support these options:

  • --action <tool>: Specify output tool (print, json, or aider)
  • --loglevel <level>: Set logging level (DEBUG, INFO, WARNING, ERROR, CRITICAL)
  • --logfile <path>: Specify log file location

Command-Specific Options

Each command has additional options that correspond to instruction templates. For example:

# Explain command options
prompts explain --files <files> --filetype <type> --user <text>

# Fix command options
prompts fix --files <files> --filetype <type> --user <text>

Examples

Print prompt for explaining Python code:

prompts explain --files main.py --filetype python

Generate JSON output for adding type hints:

prompts typehints --files utils.py --filetype python --action json

Run aider to fix Lua code:

prompts fix --files script.lua --filetype lua --user "Fix memory leak" --action aider

Ask a question using the default instructions:

prompts ask --user "What does the fix command do?"

Craft a Conventional Commit message for staged changes:

prompts commit --user "$(git diff)"

Custom Instructions

You can use custom instructions by specifying the --dir option. The custom directory must follow the same structure as the default _instructions directory. Here's how it works:

Directory Structure

custom_instructions/
├── commands/
│   ├── <command1>/
│   │   ├── command.md
│   │   ├── <key1>.md --> "Text with {value1} as placeholder."
│   │   └── <key2>/
│   │       └── <value2>.md
│   ├── <command2>/
│   │   └── command.md
│   └── ...
└── default/
    ├── <key1>.md
    └── <key2>/
        └── <value2>.md
  • commands/: Contains subdirectories for each command (e.g., explain, fix, etc.)

    • Each command directory must contain a command.md file (base instruction)
    • Additional markdown files or subdirectories correspond to command options (e.g., --files, --user, etc.). Here, the files and subdirectories handle the cli values differently:
      • File: The content of the value can be inserted in the markdown using a python placeholder. For example, for the cli option --files=foo.py, the "files.md" file contains a placeholder {files} which will be replaced by foo.py.
      • Subdirectory: The value represents a filename (without extension) of a markdown file within the subdirectory, which will be read.
  • default/: Contains fallback instructions in case an instruction is not found in the command directory.

Using a fixed directory structure allows the CLI to adapt to the custom instructions. For example, the directory structure above results in the following CLI:

prompts <command1> --key1 <value1> --key2 <value2>
prompts <command2> --key1 <value1> --key2 <value2>

Here, command2 will fallback to the default directory for options, key1 represents a file, value1 will be inserted in key1.md, key2 represents a directory and value2 is a file within the key2 directory.

Troubleshooting

If you encounter any issues, please report them on the issue tracker at: bartste-prompts issues

Contributing

Contributions are welcome! Please see CONTRIBUTING for more information.

License

Distributed under the MIT License.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

bartste_prompts-1.3.0.tar.gz (22.5 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

bartste_prompts-1.3.0-py3-none-any.whl (31.2 kB view details)

Uploaded Python 3

File details

Details for the file bartste_prompts-1.3.0.tar.gz.

File metadata

  • Download URL: bartste_prompts-1.3.0.tar.gz
  • Upload date:
  • Size: 22.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.7

File hashes

Hashes for bartste_prompts-1.3.0.tar.gz
Algorithm Hash digest
SHA256 a9c510204623141ea7a07a177ed63986c47932354f01da5d951b9de3f408af4c
MD5 4469f87056fe684f36c11812669ed7fa
BLAKE2b-256 e6785f546019f43b68f5a12f2d3039fb81daaa857906501eac7e86dc4e9be798

See more details on using hashes here.

Provenance

The following attestation bundles were made for bartste_prompts-1.3.0.tar.gz:

Publisher: release.yml on BartSte/bartste-prompts

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file bartste_prompts-1.3.0-py3-none-any.whl.

File metadata

File hashes

Hashes for bartste_prompts-1.3.0-py3-none-any.whl
Algorithm Hash digest
SHA256 c5809c2f3541206e00e4ce2bb69f58ff04a7533d0e01e20ca5fdb982ad9b6bd1
MD5 02cd3102a6d99e4c12f59ec7cd691efb
BLAKE2b-256 cbcf736acfcc488f7bf09e27ce89621f4ecdfe20d3a15ac14d1c896f2178f14c

See more details on using hashes here.

Provenance

The following attestation bundles were made for bartste_prompts-1.3.0-py3-none-any.whl:

Publisher: release.yml on BartSte/bartste-prompts

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page