Managed Agents through the Gemini API is @GoogleAI’s response to Anthropic Managed Agents

Since it’s powered by the new Antigravity agent built on Gemini 3.5 Flash, it is the most cost-effective generalized agent sandbox available to developers

We built day 1 support for Gemini Managed Agents through LlamaParse and LiteParse - this lets you give the agent access to high-quality tools for parsing PDFs.

Check it out! https://github.com/run-llama/antigravity-demo…

run-llama/antigravity-demo

Source: https://github.com/run-llama/antigravity-demo

antigravity-llamaindex

A small CLI to set up and run an Antigravity agent backed by your own GitHub repository as a data source, with LlamaParse/LiteParse skills for parsing unstructured files.

Requirements

• Python >=3.13
• uv
• gh authenticated (gh auth login)
• A Google/Gemini API key in GOOGLE_API_KEY or GEMINI_API_KEY
• (Optional) A LlamaCloud key in LLAMA_CLOUD_API_KEY or LLAMA_PARSE_API_KEY to enable LlamaParse

Environment variables can be set in a local .env file.

Install

uv tool install .

This exposes the llamagrav command.

Usage

The CLI has four commands: git-wiz, setup, run and resets. They are meant to be run in order — each one persists state in .config.json in the current directory.

1. llamagrav git-wiz — publish your data to GitHub

Initializes a git repo from a local directory, creates a GitHub repository, pushes the contents, and saves the repository URL to .config.json.

llamagrav git-wiz \
    --directory ./my-data \
    --owner my-github-user \
    --repo-name my-antigravity-data \
    --description "Data for my Antigravity agent"

Flags:

2. llamagrav setup — prepare the agent environment

Provisions the remote Antigravity environment from the repository configured above and saves the resulting environment id in .config.json.

llamagrav setup                  # uses LlamaParse (sends LlamaCloud key)
llamagrav setup --no-send-api-key   # LiteParse only, no key sent

3. llamagrav run — run the agent

llamagrav run --prompt "Summarize the PDFs in /data"

4. llamagrav reset — reset the environment

Reset the environment in .config.json to null so that you can set up a new one.

llamagrav reset

Config file

.config.json is created and updated by the CLI. It looks like:

{
  "github": { "repository_url": "https://github.com/owner/repo" },
  "environment": { "id": "...", "has_api_key": true }
}

LlamaIndex 🦙 (@llama_index): 🚀 The team at @Google just released the Agents API, a service for building and running custom agents inside a sandboxed Linux environment, and we built a template that gives these agents access to LlamaParse / LiteParse, enabling them to process unstructured documents