pricing-tracker

Pass

Audited by Gen Agent Trust Hub on Aug 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill performs legitimate data extraction and normalization of pricing tiers from publicly available SaaS and API vendor pages. It uses standard scraping patterns and focuses on structured output generation.
  • [DATA_EXPOSURE_&_EXFILTRATION]: No sensitive data access or unauthorized exfiltration patterns were detected. The network operations are limited to scraping user-provided or vendor-specific pricing URLs, which is consistent with the skill's primary purpose. It also appropriately handles 'sources' and 'capturedAt' metadata for transparency.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external pricing pages. While this presents a surface for indirect prompt injection, the risk is mitigated by instructions to normalize data into a strict JSON schema where numbers must be scalars and text is scoped to specific fields like 'notes' or 'features'. Findings in this category are low risk.
  • [COMMAND_EXECUTION]: The skill does not contain any shell commands, subprocess calls, or privilege escalation attempts. Examples provided in the documentation are for API usage patterns (agent.run) and do not execute arbitrary code.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 17, 2026, 06:28 AM
Security Audit — agent-trust-hub — pricing-tracker