Caveat verdict

agent-browser

agent-browser-openclaw

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

Encodes data AND uses eval

Same agent-browser CLI documentation variant; standard browser automation with no deceptive or malicious behavior.

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.

20
security
80
transparency
70
maintenance

What it does

These are capability combinations: each listed behavior occurs in the skill, but Caveat detects co-occurrence — it does not verify that one flows into another. Read the code to confirm a live chain.

Capability combination critical

Encodes data AND uses eval — the obfuscated-execution pattern (atob + eval; data-flow not verified)

LLM05 · ASI05 · ASI10

Findings (3)

Pattern match high

Accesses shell history/config

references/authentication.md · code · ~/.profile

Pattern match medium

Opens WebSocket connection

SKILL.md · code · WebSocket

Pattern match medium

Instructs covert action — may act without user awareness

references/snapshot-refs.md · prose · downgraded · silently

Permissions & capabilities

No declared permissions — minimal attack surface.

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