postedly-track-orders
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from the Postedly API tools, which constitutes an ingestion point for untrusted content.
- Ingestion points: Order data and delivery reports retrieved through the postedly_order_status, postedly_orders, and postedly_delivery_report tools in SKILL.md.
- Boundary markers: Instructions explicitly forbid the agent from inventing status, constructing receipts, or inferring success without evidence.
- Capability inventory: The skill is limited to reading order and delivery records via specified Postedly tools using the owner's OAuth connection.
- Sanitization: The instructions require the agent to keep meanings separate and report only recorded states, effectively sanitizing the output based on verified tool returns.
Audit Metadata