skills/lvlup-sw/axiom/harden/Gen Agent Trust Hub

harden

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for static and qualitative analysis of source code to identify resilience anti-patterns such as silent catch blocks, missing timeouts, and resource leaks. Its operations are restricted to scanning and reporting on provided code context.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes untrusted source code during the assessment process.
  • Ingestion points: Source code files provided in the user-defined scope (SKILL.md, Step 3).
  • Boundary markers: None explicitly defined in the instructions to separate code content from analysis instructions.
  • Capability inventory: Utilizes deterministic scanning tools (axiom:scan) and performs qualitative review through file reading.
  • Sanitization: No explicit sanitization or filtering of input code content is documented in the skill logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 06:38 PM
Security Audit — agent-trust-hub — harden