paw-cra-design-batch
Pass
Audited by Gen Agent Trust Hub on Apr 2, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted external data from content calendars or campaign briefs to generate AI prompts, creating a surface for indirect prompt injection.
- Ingestion points: Campaign brief and content calendar data parsed in
references/brief-intake.md(Step 1). - Boundary markers: Absent; visual directions and copy are interpolated directly into generation prompts in
references/batch-generation.md(Step 2). - Capability inventory: Network access via
curlto fal.ai; file system writes to organized campaign bundle folders in the.pawbytes/directory. - Sanitization: Absent; the skill does not explicitly filter or validate ingested data for embedded instructions or malicious patterns.
- [EXTERNAL_DOWNLOADS]: Fetches visual assets and performs generation requests via the well-known fal.ai service (
https://queue.fal.run). This activity is central to the skill's primary purpose. - [COMMAND_EXECUTION]: References the use of standard processing tools including
ffmpegfor animated carousels andPuppeteerorPlaywrightfor template rendering. These tools are verified for availability during the initialization phase.
Audit Metadata