dev-capsem-doctor

Run pytest-based diagnostic suites inside Capsem VMs to validate sandbox health.

68|10|Updated Sep 26, 2025
One-click install
npx skills add https://github.com/google/capsem --skill dev-capsem-doctor
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dev-capsem-doctor
Source: https://github.com/google/capsem/tree/main/skills/dev-capsem-doctor
Command: npx skills add https://github.com/google/capsem --skill dev-capsem-doctor

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Capsem-doctor provides end-to-end diagnostics inside the guest VM to verify sandbox integrity, network isolation, environment readiness, and AI agent functionality, acting as a gatekeeper for safe shipping.

Core Features & Use Cases

  • Pytest-based diagnostic suite that executes inside the VM to validate sandbox safety, isolation, and runtime health.
  • Covers 11 test categories and provides conftest infrastructure to support adding new tests and debugging VM sandbox issues.
  • Use cases include validating builds before shipping, diagnosing sandbox regressions, and extending test coverage for Capsem deployments.

Quick Start

Run capsem-doctor to execute the full diagnostic suite inside the Capsem VM.

Frequently Asked Questions about dev-capsem-doctor

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run pytest diagnostics to validate VM sandbox health and network isolation?

Capsem-doctor executes a pytest-based diagnostic suite inside the guest VM to validate sandbox health, network isolation, and runtime hardening. It acts as a gatekeeper to ensure Capsem deployments meet safety and reliability standards before shipping.

Can I run targeted pytest subsets instead of a full sandbox isolation test suite?

Yes, the diagnostic suite supports running targeted subsets of tests as well as the full suite. This allows you to isolate specific VM sandbox regressions or focus on individual test categories without executing the entire workflow.

What does in-VM sandbox testing cover for runtime validation and environment readiness?

In-VM sandbox testing covers 11 test categories including isolation, hardening, and runtime validation. It verifies environment readiness, sandbox integrity, and AI agent functionality to act as a gatekeeper for safe shipping.

How do I extend pytest coverage to debug new VM sandbox hardening regressions?

You can extend pytest coverage using the provided conftest infrastructure. This framework supports adding new tests and debugging existing VM sandbox issues across the 11 test categories to catch isolation and hardening regressions.

Do I need a running guest VM to validate sandbox safety and network isolation?

Yes, the diagnostic suite executes end-to-end checks directly inside the guest VM. Running the tests in-VM is required to accurately validate sandbox safety, network isolation, and runtime environment readiness.