configure-clara-receptionist

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill implements a workflow that involves reading business profile data and subsequently performing updates based on that data, which constitutes a potential surface for indirect prompt injection.\n
  • Ingestion points: Business profile and configuration data is retrieved from the Clara service via the get_business tool as specified in SKILL.md.\n
  • Boundary markers: The instructions lack specific delimiters or guardrails to treat the ingested external business data as untrusted content separate from the agent's instructions.\n
  • Capability inventory: The skill grants the agent write access to business details, greetings, voice settings, and automated SMS templates through tools like update_business and update_receptionist in SKILL.md.\n
  • Sanitization: There is no automated sanitization mentioned; however, the skill mitigates the risk by instructing the agent to show the exact proposed change and obtain user confirmation immediately before every write operation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 09:58 PM
Security Audit — agent-trust-hub — configure-clara-receptionist