skills/simota/agent-skills/chisel/Gen Agent Trust Hub

chisel

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill operates entirely on logical text analysis and transformation. It follows a rigorous Scan-Classify-Translate-Reconcile-Emit workflow that emphasizes traceability, intent preservation, and observable criteria.
  • [SAFE]: No external network operations, sensitive file path access, or suspicious shell commands were identified. All external references are relative paths to internal project documentation or standard agent modules within the shared ecosystem.
  • [SAFE]: The skill incorporates robust safety-oriented instructions, such as the 'Numeric Licensing Cascade' which prevents the fabrication of arbitrary data, and explicit prohibitions against adopting false credentials (e.g., asserting the AI is a licensed attorney).
  • [SAFE]: Guidelines for 'Ambiguity Budgeting' and 'Delete Testing' prevent over-specification and ensure that added instructions are functional and necessary, minimizing the risk of prompt bloat or unintended behavior changes.
  • [SAFE]: The skill is designed to handle potential instruction conflicts through explicit merging or precedence rules rather than silent omission, ensuring that all user constraints are accounted for or flagged for review.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 01:48 PM
Security Audit — agent-trust-hub — chisel