brutal-critic

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of purely instructional content and does not contain any executable scripts, shell commands, or tool integrations.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external content (drafts, scripts, and outlines) which represents a potential surface for indirect prompt injection. Because the skill lacks any functional capabilities like network access or file system modification, this risk is restricted to the conversation's logical flow.
  • Ingestion points: User-provided content drafts and scripts described in SKILL.md.
  • Boundary markers: None defined.
  • Capability inventory: None. No tools or automated commands are used.
  • Sanitization: None.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 11:17 PM
Security Audit — agent-trust-hub — brutal-critic