ideate

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted external data to generate content ideas.
  • Ingestion points: YouTube transcripts, newsletter issues, and web trend findings (SKILL.md, Step 1).
  • Boundary markers: Absent; the skill provides no explicit delimiters to isolate external source material from its instruction set.
  • Capability inventory: The skill utilizes subagent orchestration via the Task tool and invokes external skills such as writing:voice and branding-kit (SKILL.md, Ideation Subagents section).
  • Sanitization: Absent; there are no instructions for sanitizing or validating external input before processing.
  • [SAFE]: No hardcoded credentials, sensitive file access, or unauthorized network communication patterns were identified. The workflow relies on standard agent tool calls for web search and internal skill orchestration.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 07:59 PM