plan

Pass

Audited by Gen Agent Trust Hub on May 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its architecture, which processes untrusted external data while holding significant system privileges.
  • Ingestion points: The skill instructions encourage the use of the fetch tool for external documentation and browser-automation for web research, which can ingest malicious instructions from third-party websites into the agent's context.
  • Boundary markers: The skill lacks any instructions for the agent to use delimiters or specific safety wrappers when processing content from external sources.
  • Capability inventory: The skill utilizes powerful tools including vscode/runCommand (which can execute arbitrary IDE commands) and file system modification tools (edit/editFiles, edit/createFile, edit/createDirectory).
  • Sanitization: There is no evidence of validation or sanitization protocols for content retrieved from external repositories or web pages.
Audit Metadata
Risk Level
SAFE
Analyzed
May 17, 2026, 03:33 PM
Security Audit — agent-trust-hub — plan