us-business-registry-open-data

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill tells users/agents to "register a free Socrata app token and send it as X-App-Token," which encourages embedding an API token directly into requests/commands (e.g., headers or curl snippets) and therefore requires the LLM to handle or output secret values verbatim.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). At runtime, the skill’s fetcher (scripts/fetch_us_business_entities.py) sends LLM-readable HTTP GET requests to fixed third-party Socrata dataset resource endpoints (e.g., https://data.ny.gov/resource/n9v6-gdp6.json) and ingests the returned JSON rows/fields, so outsider-authored text can be supplied indirectly by whoever can modify those datasets.