arc-skill-sandbox
por trypto1019
Health & FitnessResumo
Test untrusted skills in an isolated environment before installing.
Descrição
O que esta skill faz
Runs untrusted skills in a monitored environment before installation. Tracks filesystem access, environment variable reads, network connections, and subprocess calls during execution. Produces a JSON report with a safety verdict (SAFE / SUSPICIOUS / DANGEROUS).
Static analysis misses runtime behavior, so executing the skill in a monitored environment reveals what it actually does with your data and credentials.
Quando usá-la
- Testing a ClawHub skill before installing it on your real agent
- Checking if a downloaded script reads API keys or tokens
- Catching outbound network calls made by an unknown skill
- Running a skill with fake credentials to detect exfiltration attempts
- Generating a safety report before a team-wide skill rollout
Exemplo de fluxo de trabalho
User asks: sandbox this skill at ~/.openclaw/skills/some-skill/ and tell me if it's safe to install
- 1
Runs sandbox.py against the skill directory with --monitor-network and --fake-env flags
- 2
Captures all filesystem operations, environment variable accesses, and outbound network requests
- 3
Injects fake API keys and tokens to observe whether the skill attempts to exfiltrate them
- 4
Generates a JSON report listing all observed activity and flags suspicious patterns
- 5
Returns the safety verdict with a summary of findings
JSON safety report covering filesystem, network, subprocess, and env var activity with a final SAFE / SUSPICIOUS / DANGEROUS verdict
Instalação
clawhub install trypto1019/arc-skill-sandboxExecute o comando da CLI ou cole o prompt no Claude Code, Cursor, Codex ou no agente de código que você usa.