agent-browser
Fail
Audited by Snyk on Feb 28, 2026
Risk Level: HIGH
Full Analysis
HIGH W007: Insecure credential handling detected in skill instructions.
- Insecure credential handling detected (high risk: 1.00). The prompt's examples and workflows instruct the agent to fill forms and run CLI commands that embed credentials directly (e.g., fill @e2 "password123" and agent-browser fill @e2 "$PASSWORD"), which requires the LLM to include secret values verbatim in generated commands and outputs.
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.90). The skill intentionally opens arbitrary external URLs and ingests page content (see SKILL.md "Navigate: agent-browser open " and commands like snapshot/get text) and the provided templates (e.g., templates/capture-workflow.sh and templates/form-automation.sh) explicitly extract and act on public webpage text/structure, so untrusted third-party pages can influence subsequent agent actions.
Audit Metadata