claude-praetorian

Pass

Audited by Gen Agent Trust Hub on Mar 17, 2026

Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
  • [NO_CODE]: The provided skill contains only documentation and tool usage patterns in SKILL.md. No Python, JavaScript, or other executable scripts are present in the package.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it restores context derived from web research and subagent outputs. 1. Ingestion points: Context is restored via 'praetorian_restore' (SKILL.md). 2. Boundary markers: No protective delimiters or sanitization instructions are provided in the skill text. 3. Capability inventory: The skill implies file system access for context storage. 4. Sanitization: None described in the documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 17, 2026, 02:03 PM