team-live-ops

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill's orchestration model creates a surface for indirect prompt injection by processing untrusted data.
  • Ingestion points: The agent ingests the [season name or event description] user argument and reads project files like design/live-ops/economy-rules.md.
  • Boundary markers: The instructions do not implement delimiters (such as XML tags) or 'ignore' instructions to isolate external data from the subagent's task context.
  • Capability inventory: Subagents are authorized to use tools including Bash, Write, Edit, and Task, allowing for file system and system operations based on the planning output.
  • Sanitization: The skill lacks logic to sanitize, validate, or escape the content of the user argument or read files before interpolation into subagent prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 12:17 AM
Security Audit — agent-trust-hub — team-live-ops