hackathon-judge-simulator

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill is entirely instructional and does not include any executable scripts, external dependencies, or command-line operations.
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it processes untrusted user-provided text. Evidence Chain: 1. Ingestion points: Multiple input fields in SKILL.md including 'problem_statement', 'solution_summary', and 'pitch_content'. 2. Boundary markers: Absent; there are no delimiters or instructions to ignore embedded commands. 3. Capability inventory: None; the skill lacks any network, file system, or subprocess execution capabilities. 4. Sanitization: Absent. The risk is considered safe because the skill does not have any functional 'sinks' (like shell access or internet connectivity) to execute or transmit malicious payloads.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 10:37 AM