hackathon-cash-prize

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No evidence of prompt injection or instructions to bypass safety guidelines was found. The instructions focus strictly on the hackathon prize domain.
  • [DATA_EXPOSURE]: No hardcoded credentials, sensitive file paths, or data exfiltration patterns were detected.
  • [OBFUSCATION]: The skill uses clear, readable markdown. No Base64 encoding, zero-width characters, or homoglyph attacks were identified.
  • [REMOTE_CODE_EXECUTION]: The skill does not perform any remote code execution or download external scripts.
  • [COMMAND_EXECUTION]: No shell commands, subprocess spawns, or privilege escalation attempts were found.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided hackathon details for decision support but lacks capabilities (such as file writing or network requests) that could be exploited via malicious data ingestion.
  • [DYNAMIC_CONTEXT_INJECTION]: No use of dynamic context injection syntax (!command) was found in the instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 09:14 PM
Security Audit — agent-trust-hub — hackathon-cash-prize