team-combat

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill takes user input via the [combat feature description] argument and passes it into the prompts of multiple subagents (game-designer, gameplay-programmer, etc.). This creates a vulnerability where a malicious user could craft a description containing instructions to override the subagents' intended behavior.
  • Ingestion points: SKILL.md (argument-hint)
  • Boundary markers: Absent (The skill does not use delimiters or instructions to ignore embedded commands in the user input)
  • Capability inventory: Bash, Task, Write, Edit, TodoWrite
  • Sanitization: Absent (No validation or filtering of the input is performed before interpolation)
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 12:16 AM
Security Audit — agent-trust-hub — team-combat