google-sheets

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs the user to paste the entire Service Account JSON and then embed that JSON verbatim into a command (ruby ... 'PASTED_JSON_CONTENT'), forcing the LLM to receive and output sensitive credentials.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The code itself is a straightforward Google Sheets client (no obfuscated payloads or remote shells) but the setup/flow explicitly instructs the user to paste a Google Service Account JSON into the assistant/chat environment, which is a high-risk credential-exfiltration pattern that would enable full programmatic access to the user's spreadsheets (and possibly other Google APIs) if abused.

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly fetches and parses arbitrary, user-provided Google Sheets content via the Google Sheets API (see SKILL.md Flow and scripts/read.rb / scripts/list_sheets.rb), which is untrusted third‑party data and is used to drive analysis and subsequent actions (including writes with user confirmation).