dont-make-me-think

Pass

Audited by Gen Agent Trust Hub on Apr 23, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted content from external URLs and user-provided code, creating a surface for indirect prompt injection. 1. Ingestion points: External websites (via /browse tool) and UI source code. 2. Boundary markers: Absent; no specific instructions to ignore embedded directives in input data. 3. Capability inventory: Permission to read and modify local source files in 'Redesign Mode'. 4. Sanitization: Absent; no filtering of ingested content is specified.
  • [SAFE]: No malicious code, unauthorized exfiltration patterns, or hardcoded credentials were detected in the skill files. The author identification is consistent with the provided context.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 23, 2026, 06:33 AM
Security Audit — agent-trust-hub — dont-make-me-think