Caveat verdict

agentbox-clawhub

agentbox-sandbox

88
🟢 Trusted
No high-risk patterns surfaced by the deep scan — automated capability review, not behavioral proof.

This is purely a documentation/guidance skill that answers questions about the AgentBox cloud sandbox product based on bundled reference docs; it does not execute code, access credentials, or make network calls — it instructs the agent to use the AgentBox SDK in standard ways.

⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.

Automated static analysis — not a human review. Caveat flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.

60
security
60
transparency
70
maintenance

Findings (2)

Pattern match critical

Possible hardcoded credential

SKILL.md · code · api_key="ab_xxxxxxxxxxxxxxxxxxxxxxxxx

Pattern match medium

Python os.environ.get — reads environment variable

references/agentbox-official.md · code · os.environ.get(

Permissions & capabilities

No declared permissions — minimal attack surface.

credential_access

Is this flag fair?

Check another skill Browse the registry Auditing your own skills or configs? Use the API