business-listings-setup
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill references a local shell script (
check_external_links.sh) for link verification logic. This script is part of the local project scaffold and is used for a functional audit task rather than executing untrusted remote code. - [INDIRECT_PROMPT_INJECTION]: The skill fetches external social profile URLs from the site's configuration to verify their existence. While fetching external data is an ingestion surface, the risk is mitigated as the skill only evaluates HTTP status codes.
- Ingestion points: Social profile URLs defined in the
SAME_ASarray withinconfig.ts(found in SKILL.md Section 3). - Boundary markers: None; however, the agent is instructed to observe response codes rather than parsing page content.
- Capability inventory: Reads local configuration files, executes a local shell script for network checks, and performs HTTP requests.
- Sanitization: The logic is strictly limited to status code comparison (e.g., checking for 404 or 5xx errors).
- [DATA_EXPOSURE]: The instructions explicitly state that sensitive identifiers like phone numbers and addresses should be obtained directly from the user rather than being stored in local configuration files, which aligns with security best practices for handling PII (Personally Identifiable Information).
Audit Metadata