pricing-page-signal
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data by fetching and reading public web pages from user-provided domains.\n
- Ingestion points: External content is ingested during 'Step 2 — Find the page, then read it' where the agent performs HTTP GET requests on domain subpaths.\n
- Boundary markers: The skill uses strict verification in 'Step 3 — Extract the record', requiring that all extracted plan names and prices appear verbatim in the source text, which limits the impact of potential malicious instructions.\n
- Capability inventory: The skill has capabilities to perform network requests (HEAD/GET) and write structured records to the workspace table. It does not have capabilities to send data to external third-party servers (exfiltration) or execute system commands.\n
- Sanitization: The skill implements a 'grep' verification step where model output is validated against the raw captured text, dropping any content that does not match exactly.
Audit Metadata