pflow-task-add

Warn

Audited by Gen Agent Trust Hub on Jun 25, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION]: The script reads the project's root .env file to determine the task file path. Access to .env files is considered a sensitive data exposure risk as they often contain project-wide credentials (API keys, secrets), though the script only uses the data locally for configuration and does not transmit it.
  • [PROMPT_INJECTION]: The skill acts as an ingestion point for indirect prompt injection by writing user-provided task titles and descriptions into a markdown file without sanitization or boundary markers. This content could potentially influence the behavior of agents that read the task list in the future.
  • [COMMAND_EXECUTION]: The skill executes the mdtodo command-line tool to add tasks. It correctly utilizes shell array expansion to handle user-supplied arguments, preventing shell command injection via the title or description fields.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 25, 2026, 07:19 PM
Security Audit — agent-trust-hub — pflow-task-add