gcphelpit — free CLI to scan Google Cloud for security, IAM, cost & reliability

gcphelpit is a free, open-source command-line tool that scans a snapshot of your Google Cloud project and finds security, IAM, cost, and reliability issues — with a plain-English fix for each finding.

It is mock-first: it reads a JSON snapshot of your resources, so it runs — and is fully testable — with zero live cloud access or credentials. A live GCP adapter can be layered on later behind the same interface.

🌐 Part of GoogleHelpit — a community hub of troubleshooting guides, tutorials, and tools for Google Cloud & Workspace. See the gcphelpit tool page.

Who is this for

• You want to audit a Google Cloud project without giving a tool live access — point it at an exported JSON snapshot instead.
• You want security, IAM, cost, and reliability covered in one pass, not four separate tools.
• You want each finding to come with a recommended fix in plain English, not just a rule ID.
• You want to gate CI/CD on GCP misconfigurations with a simple exit code.

Install & run

git clone https://github.com/EliyaS-123/gcphelpit-cli && cd gcphelpit-cli
python3 -m venv .venv && source .venv/bin/activate
pip install -e .

# Scan the bundled demo snapshot
gcphelpit scan

Or install straight from GitHub, no clone needed:

pip install git+https://github.com/EliyaS-123/gcphelpit-cli.git

You'll get a colour-coded table of findings, each with the offending resource and a recommended fix.

What it checks

gcphelpit checks lists all built-in checks. They span four categories, each finding paired with a plain-English fix:

Usage

gcphelpit scan                          # scan the bundled demo snapshot
gcphelpit scan -f my-project.json       # scan your own snapshot
gcphelpit scan --category security      # only security checks (repeatable)
gcphelpit scan --min-severity high      # only high/critical findings
gcphelpit scan --format json            # machine-readable output
gcphelpit scan --fail-on high           # exit non-zero for CI/CD gating
gcphelpit checks                        # list every check in the catalog

Exit codes: 0 clean, 1 findings at/above --fail-on, 2 usage/error.

CI/CD gating

Fail the build when a project snapshot has issues at or above a severity — for example in GitHub Actions:

# .github/workflows/gcp-audit.yml
name: GCP audit
on: [push, pull_request]
jobs:
  scan:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"
      - run: pip install git+https://github.com/EliyaS-123/gcphelpit-cli.git
      - run: gcphelpit scan -f snapshot.json --fail-on high

The snapshot

A snapshot is a plain JSON object describing what you collected from a project. Every top-level key is optional — checks simply skip data that isn't there:

{
  "project_id": "my-project",
  "buckets": [ { "name": "assets", "uniform_bucket_level_access": true, "iam_bindings": [] } ],
  "firewalls": [],
  "instances": [],
  "disks": [],
  "addresses": [],
  "service_accounts": [],
  "iam_policy": { "bindings": [] },
  "sql_instances": [],
  "budgets": []
}

See fixtures/insecure_project.json for a fully populated example (and clean_project.json for a passing one).

How it compares

gcphelpit's niche is being the scanner you can point at an exported snapshot with no credentials, covering all four categories at once with plain-English fixes. For broad multi-cloud security coverage, tools like Prowler are stronger. See the honest gcphelpit vs Prowler / ScoutSuite / gcp-auditor comparison.

Adding a check

Every check is one decorated function. Drop it in the right file under src/gcphelpit/checks/ and it auto-registers:

from ..catalog import check
from ..models import Category, Detail, ResourceRef, Severity

@check(id="SEC099", title="…", category=Category.SECURITY,
       severity=Severity.HIGH, references=["https://cloud.google.com/…"])
def my_check(snapshot):
    for bucket in snapshot.get("buckets", []):
        if bad(bucket):
            yield Detail(
                resource=ResourceRef("storage.bucket", bucket["name"]),
                message="What's wrong.",
                recommendation="How to fix it.",
            )

Development

pip install -e ".[dev]"
pytest

License

MIT