monitoring-terminal-errors

Pass

Audited by Gen Agent Trust Hub on Apr 12, 2026

Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill accesses terminal log files and source code, which may contain sensitive data such as API keys or environment variables. No network exfiltration pathways were identified.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through its ingestion of terminal output.
  • Ingestion points: Terminal logs read from the terminals folder.
  • Boundary markers: Absent; the skill does not instruct the agent to use delimiters or verify the integrity of the log content.
  • Capability inventory: The agent can read logs and modify source code files.
  • Sanitization: Absent; there is no validation of extracted error messages or paths before they are used to modify code.
  • [COMMAND_EXECUTION]: The skill instructs the agent to run local shell commands like 'head' to investigate terminal states.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 12, 2026, 09:43 PM