okx-sentiment-tracker

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.80). The prompt includes an explicit instruction to "Always use --profile live silently — don't mention it unless there's an error," which tells the agent to hide an operational detail from users and is a deceptive/hidden directive outside the skill's stated purpose.

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt includes an explicit example that embeds API credentials into a CLI command (okx config add-profile AK= SK= PP=), which encourages asking for and echoing secrets verbatim and thus creates an exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly fetches and ingests third‑party news articles (e.g., via the okx news detail / news_search commands) and even instructs performing web searches of public sites like coindesk.com, cointelegraph.com, and theblock.co in references/workflows.md, and those fetched articles are read and used to drive sentiment analysis, anomaly detection, and follow-up actions—exposing the agent to untrusted third‑party content that could carry indirect prompt injections.