add-to-global-trigger

Warn

Audited by Gen Agent Trust Hub on Jun 15, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill is designed to modify the agent's toolset by editing ~/.codex/bin/workflow.py, updating Python code logic based on user instructions. This self-modifying behavior presents a risk of code injection if the user provides malicious input.\n- [COMMAND_EXECUTION]: The skill executes multiple local commands, including python3 -m unittest, rg, and a custom tool workflow, which are used to validate and manage the local configuration.\n- [DATA_EXFILTRATION]: Hardcoded absolute file paths such as /Users/gurusharan/.codex/tests are present, which expose the developer's username and local environment details.\n- [PROMPT_INJECTION]: The skill ingests user requests for global triggers and workflows and interpolates them into system control files (AGENTS.md, workflow.py) without evident sanitization or boundary markers, creating a surface for indirect prompt injection.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 15, 2026, 09:45 AM
Security Audit — agent-trust-hub — add-to-global-trigger