google-slides-review

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The sub-agent prompts explicitly require returning verbatim slide/snapshot text as "evidence" (including byte-for-byte pattern vs filled text comparisons), so any API keys/passwords or secrets present in slides or the snapshot would be output directly by the LLM.

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

• Third-party content exposure detected (medium risk: 0.65). The required workflow ingests outsider-authored free text via the Google Slides deck content itself: the sub-agents call google_slides_get_slide / google_slides_list_slides and then read fullText/pageElements (and thumbnails via Read), which are user-provided deck text that the operating user did not author in this skill’s context, creating an indirect prompt-injection surface.