data-fetcher

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs the model to prompt the user to paste API keys and then append those secret values verbatim into a .env file (and to write them into outputs), which requires the LLM to handle and emit secret values directly — a clear exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime workflow fetches free-form CSV/JSON text from outsider-authored public web APIs (e.g., OECD SDMX requests.get(...).text and pd.read_csv(StringIO(resp.text)), plus other APIs like FRED/BLS/Yahoo Finance), which is then ingested into the agent’s execution context as readable data.