app-scaffolding
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill primarily functions as a template for generating markdown-based technical documentation. It does not contain any executable scripts, system commands, or network exfiltration patterns. The instructions in SKILL.md and its references are focused on mobile development best practices and do not request or perform privileged operations.
- [PROMPT_INJECTION]: The skill processes untrusted user input to generate a mobile app specification, presenting an indirect prompt injection surface.
- Ingestion points: User-provided inputs such as APP_NAME, APP_CATEGORY, and PRIMARY_KEYWORD defined in SKILL.md.
- Boundary markers: None; user input is interpolated directly into the scaffolding template without explicit delimiters or 'ignore' instructions.
- Capability inventory: The skill itself has no execution capabilities. As per references/corrections.md, the agent environment may have access to tools like fal.ai FLUX (images), video_generate, and music_generate, but the skill only provides configuration for these tools.
- Sanitization: No validation or escaping is performed on the user-supplied data.
- Conclusion: Given that the output is confined to text-based PRDs and architectural specs, the potential for malicious impact through indirect injection is negligible.
- [EXTERNAL_DOWNLOADS]: The skill references the 'ai-agents-from-scratch' repository and various cloud service providers (RevenueCat, TelemetryDeck, OpenRouter). These are documented as architectural recommendations and are not used to download or execute remote code at runtime.
Audit Metadata