skills/bntvllnt/agent-skills/analyze/Gen Agent Trust Hub

analyze

Pass

Audited by Gen Agent Trust Hub on Mar 25, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill consists exclusively of markdown instructions and configuration files. It does not contain any executable scripts (Python, JavaScript, shell), binaries, or hidden code blocks. No external dependencies are required or performed at runtime.
  • [COMMAND_EXECUTION]: The skill does not invoke any shell commands or system-level operations. It instructs the agent to use its native abilities to read provided file paths or perform optional web searches for context gathering as part of its normal workflow.
  • [DATA_EXFILTRATION]: There is no evidence of hardcoded credentials or instructions to send data to external servers. All processing occurs within the agent's operational context using user-supplied data for the purpose of the requested analysis.
  • [PROMPT_INJECTION]: While the skill is designed to process untrusted input (such as external files or web search results), it incorporates defensive reasoning by explicitly directing its expert perspectives to look for 'misuse' and 'adversarial hypotheses' and to rate findings by severity, which acts as a built-in mitigation against indirect prompt injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 25, 2026, 04:39 PM