aba-gemini-notebook

Pass

Audited by Gen Agent Trust Hub on Jul 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill relies on the aba command-line utility to perform browser-driven actions, including account management, notebook creation, and content querying.
  • [EXTERNAL_DOWNLOADS]: The skill documentation directs users to an external GitHub repository (github.com/SainyTK/agent-browser-app-cli) for the underlying tool and issue tracking. As this is a vendor-owned resource (authored by sainytk), it is documented as a functional dependency.
  • [PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by ingesting data from external sources. It provides commands to add content via unique HTTP/HTTPS URLs (add-urls) and read or query existing notebook content (notebook read, notebook ask), which could contain malicious instructions intended to influence the agent's output.
  • [DATA_EXFILTRATION]: The skill has the capability to read and upload local files (upload-files) and submit text to external services. However, it explicitly prohibits reading sensitive system files such as accounts.json, state.json, or browser profiles, providing a defensive guardrail against accidental credential exposure.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 28, 2026, 01:05 PM
Security Audit — agent-trust-hub — aba-gemini-notebook