AgentVault Server SDK

This package provides base classes, utilities, and tools to help developers build Agent-to-Agent (A2A) protocol compliant agents within the AgentVault ecosystem.

Overview

The core component is the BaseA2AAgent abstract class, which defines the interface agent implementations must adhere to. The SDK also provides FastAPI integration helpers (create_a2a_router) to easily expose an agent implementation as a standard A2A JSON-RPC endpoint.

Packager Utility (agentvault-sdk package)

To simplify deployment, the SDK includes a command-line utility to generate standardized Docker packaging artifacts for your agent project.

Purpose: Creates a Dockerfile, a .dockerignore file, and copies necessary configuration files (like requirements.txt and agent-card.json) into an output directory, ready for building a container image.

Usage:

agentvault-sdk package [OPTIONS]

Options:

• --output-dir DIRECTORY / -o DIRECTORY: (Required) Directory to write Dockerfile and other artifacts.
• --entrypoint TEXT: (Required) Python import path to the FastAPI app instance (e.g., my_agent.main:app).
• --python TEXT: Python version for the base image tag (e.g., 3.10, 3.11). [default: 3.11]
• --suffix TEXT: Suffix for the python base image (e.g., slim-bookworm, alpine). [default: slim-bookworm]
• --port INTEGER: Port the application will listen on inside the container. [default: 8000]
• --requirements PATH / -r PATH: Path to the requirements.txt file. If not provided, it looks for ./requirements.txt in the current directory and copies it if found. A warning is issued if the SDK dependency seems missing.
• --agent-card PATH / -c PATH: Path to the agent-card.json file. If provided, it will be copied into the output directory.
• --app-dir TEXT: Directory inside the container where the application code will reside. [default: /app]
• --help: Show this message and exit.

Example:

# Assuming your agent code is in ./src and FastAPI app is in src/my_agent/main.py
# and you have ./requirements.txt and ./agent-card.json

agentvault-sdk package \
    --output-dir ./agent_build \
    --entrypoint my_agent.main:app \
    --agent-card ./agent-card.json \
    --python 3.11

Generated Files:

Running the command creates the specified output directory (e.g., ./agent_build) containing:

• Dockerfile: A multi-stage Dockerfile optimized for Python applications.
• .dockerignore: A standard file listing patterns to exclude from the build context.
• requirements.txt (if source found/provided): A copy of your requirements file.
• agent-card.json (if --agent-card provided): A copy of your agent card file.

Building and Running the Image:

After generating the files, navigate to your project root (where your agent source code and the output directory are) and run:

# Build the image (replace 'my-agent-image:latest' with your desired tag)
docker build -t my-agent-image:latest -f ./agent_build/Dockerfile .

# Run the container
docker run -d -p 8000:8000 --name my-running-agent my-agent-image:latest

(Adjust the -p flag if you used a different port)

Future Directions

• Integration with specific deployment platforms.
• More sophisticated dependency analysis.
• Advanced configuration options for Dockerfile generation.