This release includes 3 security fixes for security teams reviewing exposed deployments.
Topics
+13 more
Affected surfaces
Summary
AI summaryCloudScope now supports Google Cloud Platform alongside Azure with new GCP tools and security hardening.
Full changelog
CloudScope now supports Google Cloud Platform alongside Azure.
GCP Provider
- Cost queries via BigQuery billing export (standard + detailed)
- Recommendations and idle resource detection via GCP Recommender API
- Untagged resources via Cloud Asset Inventory API
- Budget monitoring via Billing Budgets API
- Client-side linear regression forecasting
Multi-Provider
- All 14 shared tools accept
provider: 'azure' | 'gcp' - Default provider auto-detected from configured credentials
- Dynamic server instructions reflect connected providers
- 5 prompts accept optional
providerargument
New Tools
list_projects— list accessible GCP projectsget_cross_project_costs— cost breakdown across GCP projects
Security
- Raw SDK error messages no longer returned to clients
- BigQuery table name validated against SQL injection
- ESLint rules enforce
console.erroronly (noconsole.log) and banerror.messagein logs
Testing
- 280 unit tests, 29 Azure E2E tests, 18 GCP E2E tests
npm run test:e2e/test:e2e:gcp/test:e2e:all
Setup
Azure works with az login (zero config). GCP requires BigQuery billing export — see
README for setup.
Security Fixes
- Raw SDK error messages are no longer returned to clients.
- BigQuery table name validation added to prevent SQL injection.
- ESLint configuration enforces use of `console.error` only (bans `console.log`) and disallows logging `error.message`.
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About alexpota/cloudscope-mcp
Azure cloud cost management — spending analysis, forecasts, anomaly detection, budgets, optimization recommendations, idle resource detection, tag-based cost allocation, and cross-subscription queries through natural language.
Related context
Beta — feedback welcome: [email protected]