orchestrator-emergency

Pass

Audited by Gen Agent Trust Hub on Apr 2, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes external, untrusted incident data.
  • Ingestion points: The skill is designed to analyze and summarize user-provided or system-generated outage reports and log data (SKILL.md).
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to disregard embedded commands or formatting within the incident data.
  • Capability inventory: The agent is instructed to suggest using the memory-recorder tool to persist incident details to Serena memory, creating a path for long-term poisoning.
  • Sanitization: There are no instructions or mechanisms for validating, escaping, or filtering the input data before it is summarized or recorded.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 2, 2026, 07:25 PM
Security Audit — agent-trust-hub — orchestrator-emergency