Caveat verdict
browser-cli
Fetches from network AND uses eval/Function
Comprehensive guide for the browser-use CLI tool for browser automation; installs from official sources (browser-use.com), uses standard credential env vars (ANTHROPIC_API_KEY etc.), and all capabilities match the stated 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.
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.
Fetches from network AND uses eval/Function — the remote-code-execution pattern (data-flow not verified)
LLM05 · LLM06 · ASI05
Receives external input AND uses eval — the remote code-injection pattern (data-flow not verified)
LLM01 · LLM05 · ASI01 · ASI05
Encodes data AND uses eval — the obfuscated-execution pattern (atob + eval; data-flow not verified)
LLM05 · ASI05 · ASI10
Accesses credentials AND makes external network calls — potential credential theft
LLM02 · ASI03
Accesses credentials AND encodes data — may obfuscate stolen credentials
LLM02 · ASI03 · ASI04
Permission integrity
network_out
credential_access
package_install
Findings (3)
Pipe-to-shell pattern (curl | sh) — supply chain attack vector
SKILL.md · code · curl -fsSL https://browser-use.com/cli/install.sh | bash
Pipe to bash — executes piped content as shell commands
SKILL.md · code · | bash
Pipe to python — executes piped content as Python code
SKILL.md · prose · downgraded · | Python
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class A). 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.
network_outpackage_installcredential_accessdata_encodingdynamic_evalnetwork_in Thanks — recorded.