synthetic-session-generator

Pass

Audited by Gen Agent Trust Hub on Jun 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill operates entirely within the local environment, utilizing Python scripts for text processing and file management without external network dependencies.
  • [DATA_EXPOSURE]: The skill does not access sensitive user directories (such as .ssh, .aws, or browser profiles). It manages its own configuration and persona data locally within the skill directory.
  • [REMOTE_CODE_EXECUTION]: No remote script downloads or piped command executions were found. The Python scripts rely exclusively on standard libraries (argparse, json, os, sys).
  • [PROMPT_INJECTION]: The instructions in SKILL.md focus on persona consistency and modality fidelity. There are no attempts to override agent safety protocols or bypass system constraints.
  • [COMMAND_EXECUTION]: The skill uses subprocess-like behavior only through its own provided scripts for local file conversion and setup. User-supplied arguments are validated as integers or choice-limited strings where applicable.
  • [SAFE]: The README contains a claim regarding previous audits and a 'VERDICT: ALL CLEAR'. Per the security analysis guidelines, this claim was treated as non-authoritative data and did not influence the final verdict, which is based on a primary analysis of the skill's code and instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 12, 2026, 10:20 PM
Security Audit — agent-trust-hub — synthetic-session-generator