eo-project-update

Pass

Audited by Gen Agent Trust Hub on May 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for project management tasks, including updating task statuses in phases and recording project decisions.
  • [COMMAND_EXECUTION]: The skill instructions do not contain any shell commands, subprocess calls, or system-level execution patterns. It relies entirely on structured file updates.
  • [DATA_EXFILTRATION]: There are no network operations, API calls, or external data transfers. The skill only interacts with local files defined in the project configuration.
  • [PROMPT_INJECTION]: The skill uses a defined mapping of user input to specific actions (e.g., '完成' maps to task completion), which prevents arbitrary instruction execution from user prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
May 24, 2026, 04:05 AM
Security Audit — agent-trust-hub — eo-project-update