pi-agy-acp0.1.2
Google Antigravity ACP provider extension for Pi Coding Agent
Untested on omp
This is a pi package. No one has reported whether it works on omp yet. If you try it, add your result below.
Reports 0
No reports yet.
Readme
pi-agy-acp
Use Google Antigravity in Pi through the official, locally installed agy CLI.
This extension registers the antigravity-acp provider and lets Pi execute the
model's tool calls. It supports streaming text and reasoning, images, and
parallel tool calls.
Prerequisites
- Node.js >= 22.18.0.
- Pi (
@earendil-works/pi-coding-agent) 1.5.x. - The official
agyCLI >= 1.4.12, signed in with your Google account. Runagyin your terminal to complete sign-in before using the extension.
Install
Install from GitHub (recommended) or npm:
pi install git:github.com/fdkevin0/pi-agy-acp
# Or:
pi install npm:pi-agy-acp
Add -l to install only for the current project (.pi/settings.json). To try,
update, or remove the extension:
pi -e git:github.com/fdkevin0/pi-agy-acp
pi update --extensions
pi remove git:github.com/fdkevin0/pi-agy-acp
For npm, remove npm:pi-agy-acp instead. Add -l when removing a project-local
installation. See development setup for local testing.
Usage
Start pi, then use /model to select a model under antigravity-acp.
Examples include gemini-4-argon, gemini-3.8-flash, claude-opus-5-5,
claude-sonnet-5-5, and gpt-oss-120b-medium. The available list refreshes
automatically from your account.
| Command | Purpose |
|---|---|
/agy-acp or /agy-acp status |
Show runtime state, account details, models, sessions, and the last turn. |
/agy-acp quota |
Show account details, grouped quota bars, and reset countdowns. |
/agy-acp reset |
Close ACP sessions; the next request rebuilds the connection. |
After an Antigravity turn settles, the status bar shows remaining quota (for example,
AGY 80%) and warns when quota is low; see quota status.
Default prices are zero, not a promise of free calls. Pi pricing overrides are used when configured.
More information
- Reference: commands, configuration, stored data, model discovery, and limitations.
- Development: setup, request flow, source layout, and verification.
- Changelog: release history and upcoming changes.
- Attribution and MIT license.