# GCP Infrastructure

> Google Cloud Platform infrastructure inspection. Use when investigating GCP Compute instances, GKE clusters, Cloud Functions, Cloud SQL, or project metadata.

- Skill: `incidentfox/gcp-infrastructure` (Agent Skill, multi-file: 7 files)
- Install (CLI): `npx skillmds@latest add incidentfox/gcp-infrastructure`
- Raw SKILL.md: https://api.skillmd.com/api/skills/incidentfox/gcp-infrastructure/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- Author: incidentfox (https://skillmd.com/u/incidentfox)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/incidentfox/gcp-infrastructure

---


# GCP Infrastructure

## Authentication

**IMPORTANT**: Credentials are injected automatically by a proxy layer. Do NOT check for `GCP_SERVICE_ACCOUNT_KEY` in environment variables - it won't be visible to you. Just run the scripts directly; authentication is handled transparently.

Configuration environment variables you CAN check (non-secret):
- `GCP_PROJECT_ID` - GCP project ID

---

## Available Scripts

All scripts are in `.claude/skills/infrastructure-gcp/scripts/`

### list_compute_instances.py - List Compute Engine VMs
```bash
python .claude/skills/infrastructure-gcp/scripts/list_compute_instances.py [--zone us-central1-a]
```

### list_gke_clusters.py - List GKE Clusters
```bash
python .claude/skills/infrastructure-gcp/scripts/list_gke_clusters.py
```

### list_cloud_functions.py - List Cloud Functions
```bash
python .claude/skills/infrastructure-gcp/scripts/list_cloud_functions.py
```

### list_cloud_sql.py - List Cloud SQL Instances
```bash
python .claude/skills/infrastructure-gcp/scripts/list_cloud_sql.py
```

### get_project_metadata.py - Project Info
```bash
python .claude/skills/infrastructure-gcp/scripts/get_project_metadata.py
```

---

## Investigation Workflow

### GKE Cluster Issue
```
1. list_gke_clusters.py
2. Use infrastructure-kubernetes skill for pod-level debugging
```

### Compute Instance Issue
```
1. list_compute_instances.py --zone <zone>
2. Check instance status and IPs
```