GEMINI_API_KEY=<paste_your_api_key_here>

conda create --name gemini_agents python=3.12 -y

conda activate gemini_agents

pip install google-genai requests python-dotenv

from dotenv import load_dotenv
from google import genai

# Load secure environment variables
load_dotenv()

# Initialize the GenAI Client
client = genai.Client()

# Create a basic interaction with a managed agent
interaction = client.interactions.create(
    agent="antigravity-preview-05-2026",
    input="Install the matplotlib package, verify its version, and report back.",
    environment="remote"
)

# Output the status of the agent
print(f"Status: {interaction.status}")
print(f"Environment ID: {interaction.environment_id}")
print(f"Output:\n{interaction.output_text}")

Status: completed
Environment ID: 104ad7f8-32e0-4b8d-b344-24d92eb74eb6
Output:
I have successfully installed the matplotlib package in the sandbox environment and verified its installation.

Here are the details:
- **Installation Command:** python3 -m pip install --break-system-packages matplotlib
- **Installed Version:** 3.10.9

# First interaction
inter1 = client.interactions.create(
    agent="antigravity-preview-05-2026",
    input="Write a Python script sum.py that adds all integers from 1 to 100.",
    environment="remote"
)

# Second interaction
inter2 = client.interactions.create(
    agent="antigravity-preview-05-2026",
    previous_interaction_id=inter1.id, # Passes the conversation history
    environment=inter1.environment_id, # Keeps the same filesystem state
    input="Execute 'sum.py' using Python and display the standard output."
)

# Output the status of the agent
print(f"Output:\n{inter2.output_text}")

inter = client.interactions.create(
    agent="antigravity-preview-05-2026",
    input="Add all the numbers in the /workspace/numbers.txt file.",
    environment={
        "type": "remote",
        "sources": [
            {
                "type": "inline",
                # The file where to store the data in the agent environment
                "target": "/workspace/numbers.txt",
                # Assumes that the file data/numbers.txt exists
                "content": utils.read_text_file("data/numbers.txt")
            }
        ]
    }
)

inter = client.interactions.create(
    agent="antigravity-preview-05-2026",
    input="Add all the numbers in the /workspace/repository/numbers.txt file.",
    environment={
        "type": "remote",
        "sources": [
            {
                "type": "repository",
                "source": "https://github.com/fran-aubry/gemini-agents-tutorial",
                "target": "/workspace/repository"
            }
        ]
    }
)

/workspace/repository

https://generativelanguage.googleapis.com/v1beta/files/environment-<env_id>:download

def download_env(env_id, path="environments"):
    download_url = f"https://generativelanguage.googleapis.com/v1beta/files/environment-{env_id}:download"
    try:
        request_params = {"alt": "media"}  # Retrieves raw media binary
        request_headers = {"x-goog-api-key": os.environ.get("GEMINI_API_KEY")}
        # Download the environment
        print(f"Downloading environment: {env_id}")
        response = requests.get(
            download_url,
            params=request_params,
            headers=request_headers,
            allow_redirects=True
        )
        response.raise_for_status()
        # Save the compressed workspace archive locally
        archive_name = f"{env_id}.tar"
        output_path = os.path.join(path, archive_name)
        with open(output_path, "wb") as archive_file:
            archive_file.write(response.content)
        print(f"Successfully downloaded workspace snapshot archive: {output_path}")     
    except requests.exceptions.RequestException as error:
        print(f"Failed to download sandbox workspace via HTTP request: {error}")
    except tarfile.TarError as archive_error:
        print(f"Failed to unpack download tarball: {archive_error}")

agent = client.agents.create(
            id="data-analyst",
            base_agent="antigravity-preview-05-2026",
            base_environment={
                "type": "remote",
                "sources": [
                    {
                        "type": "inline", 
                        "target": ".agents/AGENTS.md", 
                        "content": read_text_file(".agents/AGENTS.md")
                    },
                    # Explicitly load the skill
                    {
                        "type": "inline", 
                        "target": ".agents/skills/csv-aggregator/SKILL.md", 
                        "content": read_text_file(".agents/skills/csv-aggregator/SKILL.md")
                    },	
                    {
                        "type": "repository",
                        "source": "https://github.com/fran-aubry/gemini-agents-tutorial",
                        "target": "/workspace/repository"
                    }
                ]
            }
)

.agents/AGENTS.md

.agents/skills/<skill_name>/SKILL.md

---
name: <skill_name>
description: <description of when to use the skill>
---
<steps on how to perform the task>

from dotenv import load_dotenv
from google import genai
import utils

load_dotenv()
client = genai.Client()

data_analyst = utils.load_or_create_agent(client, "data-analyst")
print(f"Agent '{data_analyst.id}' initialized.")

inter1 = client.interactions.create(
    agent=data_analyst.id,
    input="Install the matplotlib package.",
    environment="remote"
)

inter2 = client.interactions.create(
    agent=data_analyst.id,
    input="Use the csv-aggregator to plot the top 10 genres from /workspace/repository/data/netflix.csv in terms of viewership",
    environment=inter1.environment_id
)

inter3 = client.interactions.create(
    agent=data_analyst.id,
    input="Execute the genres.py script using python.",
    environment=inter2.environment_id
)

utils.download_env(inter3.environment_id)