zoho

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The prompt explicitly shows curl commands and setup steps that require pasting client_id, client_secret, authorization codes, and refresh_token values (and instructs creating a .env with those secrets), meaning an agent would need to handle and potentially output these secret values verbatim.

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

• Third-party content exposure detected (high risk: 0.80). The skill's standup pipeline (scripts/standup-summarizer.sh) and SKILL.md explicitly fetch meeting recordings from the Zoho Meeting API (downloadUrl/publicDownloadUrl), transcribe and summarize user-generated audio/transcripts, and those transcripts are consumed by the agent to produce summaries and drive actions (e.g., creating tasks), so arbitrary third-party user content can materially influence agent behavior.