cali-degustia-metricas

Pass

Audited by Gen Agent Trust Hub on Jul 24, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes untrusted interview transcripts that could contain embedded instructions intended to manipulate the JSON output or subsequent agent behavior.
  • Ingestion points: Processes transcript de entrevista provided as input.
  • Capability inventory: The skill writes JSON files to the local filesystem and performs web searches for wage data.
  • Boundary markers: No specific delimiters or warnings are used to instruct the agent to ignore instructions embedded within the transcript text.
  • Sanitization: No input validation or sanitization of the transcript content is specified before processing.
  • [COMMAND_EXECUTION]: The skill performs filesystem operations by creating directories and writing files (degustia-metricas/YYYY-MM-dd-cliente-metricas.json). While these are intended features for data persistence, they represent a capability to modify the local environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 24, 2026, 06:40 PM
Security Audit — agent-trust-hub — cali-degustia-metricas