review-clara-inbox

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and summarize untrusted data from external sources such as customer inquiries and call transcripts.
  • Ingestion points: Untrusted content enters the context via the list_inquiries, get_inquiry, and search tools defined in SKILL.md.
  • Boundary markers: There are no instructions to use specific delimiters or boundary markers to isolate external text from the agent's instructions.
  • Capability inventory: The skill possesses read-only access to PII, communication history, and call recordings.
  • Sanitization: No sanitization or escaping of external content is specified before the data is processed by the agent.
  • [EXTERNAL_DOWNLOADS]: The skill connects to https://mcp.heyitsclara.com/mcp to perform authentication and retrieve account data. This is the official domain for the Clara service integration.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 09:57 PM
Security Audit — agent-trust-hub — review-clara-inbox