oebb-scotty

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt embeds a hard-coded auth value ("aid":"OWDL4fE4ixNiPBBm") and explicitly shows/requests it be included verbatim in every API request, meaning the LLM must handle and output a credential-like secret.

HIGH W008: Secret detected in skill content (API keys, tokens, passwords).

• Secret detected (high risk: 1.00). I scanned the entire skill prompt for literal, high-entropy values used as credentials. The JSON auth block repeatedly contains: "auth":{"type":"AID","aid":"OWDL4fE4ixNiPBBm"}

This 16-character mixed-case alphanumeric value is not a placeholder (e.g., YOUR_API_KEY or sk-xxxx), is not truncated/redacted, and is used directly in the auth field of example requests — all indicators of a real, embedded credential. Station IDs, dates, times, product masks, and client id/name strings are low-entropy or intended identifiers and are not secrets, so they were ignored. No private keys, JWTs, or other high-entropy secrets were found beyond the AID above.