hackathon-mentoring

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [NO_CODE]: The skill provides instructional content for hackathon organization and does not include any scripts, executable binaries, or system-level commands.
  • [INDIRECT_PROMPT_INJECTION]: The skill includes an interactive interview process that ingests user input to customize the mentoring plan.
  • Ingestion points: The 'Interview' section in SKILL.md prompts for user answers to 10 specific questions.
  • Boundary markers: The instructions recommend using multiple-choice questions to limit input variability.
  • Capability inventory: None. The skill does not possess the ability to execute shell commands, perform network requests, or modify the local filesystem.
  • Sanitization: No explicit sanitization or input validation is defined.
  • Conclusion: As the skill has no dangerous capabilities, the risk of exploiting user input is negligible.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 09:14 PM
Security Audit — agent-trust-hub — hackathon-mentoring