shipping-cost

Pass

Audited by Gen Agent Trust Hub on Mar 23, 2026

Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill includes the instruction "Do NOT ask the user questions," which overrides standard agent interaction protocols and increases the risk of autonomous action without user confirmation.
  • [DATA_EXFILTRATION]: The skill attempts to access and modify files within ~/.claude/projects/ for telemetry purposes. This is a sensitive path containing internal tool configuration and project history.
  • [PROMPT_INJECTION]: The skill processes untrusted external data (carrier rates, invoices, and codebase) without implementing boundary markers or instructions to ignore embedded commands, creating a surface for indirect prompt injection.
  • Ingestion points: Reads codebase, carrier rate tables, and invoices.
  • Boundary markers: Absent. No delimiters or warnings are used when processing external data.
  • Capability inventory: File system read, directory creation (docs/), and file writing (docs/shipping-cost-analysis.md).
  • Sanitization: Absent. No sanitization of ingested content before processing is specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 23, 2026, 10:57 AM
Security Audit — agent-trust-hub — shipping-cost