competition-submission-checker

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and act upon data from untrusted or external project files which may contain embedded instructions.
  • Ingestion points: The agent is directed to read instructions and context from AGENTS.md, docs/memory-bank/progress.md, and project README files.
  • Boundary markers: There are no instructions to treat content from these external files as untrusted or to ignore instructions embedded within them.
  • Capability inventory: The skill does not provide its own scripts or commands, relying instead on the agent's inherent file-reading capabilities.
  • Sanitization: No sanitization or validation logic is defined for the content read from external documentation.
  • [NO_CODE]: The skill contains only documentation and instruction files (SKILL.md and CHANGELOG.md). No scripts, executables, or dynamic execution patterns were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 03:51 AM
Security Audit — agent-trust-hub — competition-submission-checker