@nilskluewer/pi-anthropic-vertex-enterprise-agent-platform
Anthropic Claude models on Google Cloud Vertex AI / Gemini Enterprise Agent Platform for pi.
Package details
Install @nilskluewer/pi-anthropic-vertex-enterprise-agent-platform from npm and Pi will load the resources declared by the package manifest.
$ pi install npm:@nilskluewer/pi-anthropic-vertex-enterprise-agent-platform- Package
@nilskluewer/pi-anthropic-vertex-enterprise-agent-platform- Version
0.1.0- Published
- Jul 9, 2026
- Downloads
- 27/mo · 27/wk
- Author
- nilskluewer
- License
- MIT
- Types
- extension
- Size
- 20.8 KB
- Dependencies
- 1 dependency · 2 peers
Pi manifest JSON
{
"extensions": [
"./extensions/anthropic-vertex-enterprise-agent-platform.ts"
]
}Security note
Pi packages can execute code and influence agent behavior. Review the source before installing third-party packages.
README
@nilskluewer/pi-anthropic-vertex-enterprise-agent-platform
Overview
Use Anthropic Claude models through Google Cloud Vertex AI / Gemini Enterprise Agent Platform in pi.
The extension creates an AnthropicVertex client and delegates streaming to Pi's
built-in anthropic-messages implementation, so Pi still handles tools, prompt
caching, thinking blocks, partial JSON streaming, usage tracking, and cost calculation.
Getting Started
pi install npm:@nilskluewer/pi-anthropic-vertex-enterprise-agent-platform
pi
/setup-vertexai your-project-id eu
/model anthropic-vertex/claude-sonnet-5
Structure
.
├── extensions/
│ └── anthropic-vertex-enterprise-agent-platform.ts # Pi extension entry point
├── LICENSE
├── README.md
└── package.json # npm and Pi package manifest
Setup
Requirements:
- Google Cloud project with the Vertex AI API enabled.
- Claude models enabled in Model Garden / Gemini Enterprise Agent Platform.
gcloudCLI installed for Google Application Default Credentials.
Run the Pi setup command after installing the package:
/setup-vertexai <google-cloud-project> [location]
Example:
/setup-vertexai my-project-id eu
The command stores the project and location in Pi's user config directory and checks Google Application Default Credentials. If credentials are missing in interactive mode, it can run:
gcloud auth application-default login
Environment variables override the saved setup:
| Name | Description | Default |
|---|---|---|
GOOGLE_CLOUD_PROJECT |
Google Cloud project ID used by Vertex AI. | Saved setup value |
ANTHROPIC_VERTEX_PROJECT_ID |
Claude Code compatible fallback project ID. | Optional |
GCLOUD_PROJECT |
Fallback project ID if the values above are not set. | Optional |
GOOGLE_CLOUD_LOCATION |
Vertex AI region or multi-region. Use eu for the EU multi-region endpoint. |
Saved setup value or eu |
CLOUD_ML_REGION |
Claude Code compatible fallback region. | Optional |
VERTEX_REGION |
Additional fallback region. | Optional |
Run
Try the package without adding it to your settings:
pi -e npm:@nilskluewer/pi-anthropic-vertex-enterprise-agent-platform
Use any Claude model exposed by Pi's built-in Anthropic model list and enabled in your Vertex AI project. These models are verified with this package:
pi --provider anthropic-vertex --model claude-fable-5
pi --provider anthropic-vertex --model claude-sonnet-5
pi --provider anthropic-vertex --model claude-opus-4-8
The extension registers Claude model metadata from Pi's built-in Anthropic provider at runtime. If Pi adds newer Claude models and they are enabled in your Vertex AI project, they can be used through this provider as well.
For the eu and us multi-regions, Vertex AI uses aiplatform.<region>.rep.googleapis.com
endpoints. The package defaults to eu; set another location in /setup-vertexai or
GOOGLE_CLOUD_LOCATION when you need a specific supported region such as us-east5.
License
MIT