yc-office-hours
Pass
Audited by Gen Agent Trust Hub on Mar 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates purely as a conversational framework for startup diagnostics and does not include any executable code or scripts.
- [SAFE]: It contains a 'HARD GATE' instruction explicitly prohibiting the agent from writing code, scaffolding projects, or taking any implementation actions, which mitigates risks of unintended command execution.
- [SAFE]: References to external domains are limited to 'ycombinator.com', which is a well-known and trusted service in the technology industry.
- [SAFE]: No evidence of prompt injection, obfuscation, data exfiltration, or persistence mechanisms was detected in the instructions or metadata.
Audit Metadata