skill-deck
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user-supplied briefs and external research results which could influence the agent's behavior or output content. -- Ingestion points: User-provided topics and descriptions (SKILL.md, Step 1) and external research gathered via subagent tools (SKILL.md, Step 2). -- Boundary markers: None explicitly defined to separate untrusted content from system instructions. -- Capability inventory: The skill uses the document-skills:pptx tool to write files to the local file system (SKILL.md, Step 4). -- Sanitization: The skill implements a mandatory 'wireframe gate' (SKILL.md, Step 3) that requires the user to manually approve the generated outline before the final PPTX is rendered, serving as a significant security checkpoint.
- [EXTERNAL_DOWNLOADS]: The skill references a requirement for the document-skills plugin and provides instructions for its installation from a trusted source (anthropic-agent-skills).
Audit Metadata