Caveat verdict

agb(agent-browser)

agb

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

Receives external input AND uses eval

Browser automation CLI wrapper with documented commands for navigation, snapshots, and interaction; behavior matches the stated web automation purpose.

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.

10
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

Receives external input AND uses eval — the remote code-injection pattern (data-flow not verified)

LLM01 · LLM05 · ASI01 · ASI05

Findings (4)

Pattern match high

Uses eval() — can execute arbitrary code

index.js · prose · downgraded · eval(

Pattern match medium

Opens WebSocket connection

SKILL.md · code · WebSocket

Pattern match medium

References child_process — can spawn system processes

index.js · prose · downgraded · child_process

Pattern match medium

Uses spawn() — can execute external programs

index.js · prose · downgraded · spawn(

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class D). Final tier capped at Caution — cannot be lifted by any downgrade, example-payload opt-in, or allowlist.

Permissions & capabilities

No declared permissions — minimal attack surface.

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