a0-browser-ext

Pass

Audited by Gen Agent Trust Hub on May 4, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides a structured and secure framework for creating and installing browser extensions. It explicitly mandates that the agent uses a 'Safety First' approach, which includes stating requested behaviors clearly, limiting host permissions, and avoiding remote code execution or credential collection.
  • [PROMPT_INJECTION]: The skill contains a surface for indirect prompt injection because it is designed to ingest and summarize data from external Chrome extensions.
  • Ingestion points: The agent parses manifest.json files and extension content from the Chrome Web Store (SKILL.md).
  • Boundary markers: No specific delimiters are defined to separate untrusted extension data from the agent's internal reasoning during the inspection phase.
  • Capability inventory: The skill is capable of writing files to the local filesystem (under /a0/usr/browser-extensions/), extracting ZIP archives, and restarting browser runtimes using Playwright (SKILL.md).
  • Sanitization: The skill relies on manual inspection and auditing of permissions and source code by the agent and user to identify and reject suspicious capabilities before installation.
Audit Metadata
Risk Level
SAFE
Analyzed
May 4, 2026, 03:40 PM