ayaiay 1.0.1

CLI and SDK for the AyAiAy.org AI agents marketplace

AyAiAy CLI

CLI and SDK for the AyAiAy.org AI agents marketplace - discover, install, and manage AI Agent Profiles, Instruction Packs, and Prompt Packs.

Installation

pip install ayaiay

For development:

pip install -e ".[dev]"

Quick Start

Search for Packs

# Search by keyword
ayaiay search "code review"

# Filter by type
ayaiay search --type agent
ayaiay search --type instruction
ayaiay search --type prompt

# Filter by tags
ayaiay search --tag python --tag testing

Install Packs

# Install latest version
ayaiay install acme/code-reviewer

# Install specific version
ayaiay install acme/code-reviewer@1.0.0

# Force reinstall
ayaiay install acme/code-reviewer --force

Package Management with ayaiay.json

Similar to composer.json in PHP or package.json in Node.js, you can use ayaiay.json to track all your project's pack dependencies with their versions.

# Initialize a new ayaiay.json file
ayaiay init

# Add a pack to ayaiay.json and install it
ayaiay add acme/code-reviewer
ayaiay add acme/code-reviewer@1.0.0

# Remove a pack from ayaiay.json and uninstall it
ayaiay remove acme/code-reviewer

# Sync installed packs with ayaiay.json (useful after cloning a project)
ayaiay sync

# Update all packs to their latest versions
ayaiay update

# Update a specific pack
ayaiay update acme/code-reviewer

Example ayaiay.json:

{
  "version": "1.0",
  "packages": {
    "acme/code-reviewer": {
      "name": "acme/code-reviewer",
      "version": "1.2.0",
      "installed_at": "2024-01-15T10:30:00Z",
      "digest": "sha256:abc123...",
      "dependencies": {}
    }
  },
  "updated_at": "2024-01-15T10:30:00Z"
}

Manage Installed Packs

# List installed packs
ayaiay list

# Show pack details
ayaiay show acme/code-reviewer

# Uninstall a pack
ayaiay uninstall acme/code-reviewer

Validate Manifests

# Validate an ayaiay.yaml manifest
ayaiay validate ayaiay.yaml
ayaiay validate ./my-pack/ayaiay.yaml

CLI Configuration

# Show current configuration and API status
ayaiay info

SDK Usage

from ayaiay import AyAiAyClient, PackageManager, validate_manifest

# Search for packs
with AyAiAyClient() as client:
    results = client.search_packs(query="code review", pack_type="agent")
    for pack in results.packs:
        print(f"{pack.full_name}: {pack.description}")

# Get pack details
with AyAiAyClient() as client:
    pack = client.get_pack("acme/code-reviewer")
    versions = client.get_pack_versions("acme/code-reviewer")

# Package management with ayaiay.json
pm = PackageManager()

# Initialize lock file
pm.init()

# Add packages
success, message, result = pm.add_package("acme/code-reviewer@1.0.0")
if success:
    print(f"Added: {message}")

# List packages in lock file
packages = pm.list_packages()
for name, version, installed_at in packages:
    print(f"{name}@{version}")

# Sync packages (install missing ones)
results = pm.sync()
for package_name, success, message in results:
    print(f"{package_name}: {message}")

# Update packages to latest versions
results = pm.update()
for package_name, success, message in results:
    print(f"{package_name}: {message}")

# Validate a manifest file
result = validate_manifest("ayaiay.yaml")
if result.is_valid:
    print(f"Valid manifest: {result.manifest.name}")
else:
    for error in result.errors:
        print(f"Error: {error}")

Configuration

Configuration is loaded from (in order of priority):

1. Environment variables (highest priority)
2. Config file (~/.ayaiay/config.yaml)
3. Default values

Environment Variables

Variable	Description	Default
AYAIAY_API_URL	API base URL	https://api.ayaiay.org
AYAIAY_REGISTRY_URL	OCI registry URL	ghcr.io/ayaiayorg
AYAIAY_INSTALL_DIR	Pack installation directory	~/.ayaiay/packs
AYAIAY_TOKEN	Authentication token	-

Config File

# ~/.ayaiay/config.yaml
api_base_url: https://api.ayaiay.org
registry_url: ghcr.io/ayaiayorg
install_dir: ~/.ayaiay/packs
cache_dir: ~/.ayaiay/cache
timeout: 30.0

Creating Packs

To publish your own packs, create an ayaiay.yaml manifest in your repository:

version: "1.0"
name: my-awesome-pack
description: An awesome AI pack for code review
author: Your Name
license: MIT
repository: https://github.com/you/my-awesome-pack
tags:
  - code-review
  - python

agents:
  - name: code-reviewer
    description: Reviews code for quality and best practices
    system_prompt: |
      You are an expert code reviewer. Analyze code for:
      - Best practices
      - Performance issues
      - Security vulnerabilities
    model: gpt-4
    tools:
      - read_file
      - write_file

instructions:
  - name: coding-standards
    description: Coding standards to follow
    content: |
      Follow PEP 8 for Python code.
      Use meaningful variable names.
      Write docstrings for all functions.

prompts:
  - name: review-request
    description: Template for requesting code review
    template: |
      Please review the following {language} code:

      ```{language}
      {code}
      ```

      Focus on: {focus_areas}
    variables:
      - language
      - code
      - focus_areas

dependencies:
  base-pack: "^1.0.0"

Development

Setup

git clone https://github.com/ayaiayorg/ayaiay-cli.git
cd ayaiay-cli
pip install -e ".[dev]"

Running Tests

pytest
pytest --cov=ayaiay

Code Quality

# Format code
black .
isort .

# Type checking
mypy .

# Linting
ruff check .

Publishing to PyPI

This package uses GitHub Actions with Trusted Publisher (OIDC) for secure, tokenless publishing.

Setup (one-time):

1. Go to PyPI and create an account
2. Navigate to your account's "Publishing" settings
3. Add a new Trusted Publisher with:
   • Owner: ayaiayorg
   • Repository: ayaiay-cli
   • Workflow: publish.yml
   • Environment: pypi
4. Repeat for TestPyPI with environment: testpypi

Publishing:

• Automatic: Create a GitHub Release to publish to PyPI
• Manual (TestPyPI): Run the "Publish to PyPI" workflow manually, selecting testpypi
• Manual (PyPI): Run the workflow manually, selecting pypi

Before releasing:

1. Update the version in pyproject.toml
2. Commit and push changes
3. Create a GitHub Release with a tag matching the version (e.g., v0.1.0)

Automated Agent Workflow

This repository includes a GitHub Actions workflow that automatically processes issues and creates pull requests to resolve them. To use it:

1. Create an issue or use an existing one
2. Add the agent-task or automation label
3. The workflow will automatically create a branch, make changes, and open a PR

For detailed documentation, see docs/AGENT_WORKFLOW.md.

License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.

Links

• AyAiAy.org - Marketplace website
• Documentation - Full documentation
• Issues - Bug reports and feature requests
• Agent Workflow Guide - Automated issue handling