skills/gdewitte/agentic/spec/Gen Agent Trust Hub

spec

Pass

Audited by Gen Agent Trust Hub on Jul 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill facilitates the creation of technical documentation, which is a standard development task with no inherent security risks.
  • [SAFE]: It enforces a human-in-the-loop workflow, requiring explicit user approval after the spec is written and before the agent can move to implementation.
  • [COMMAND_EXECUTION]: The skill manages local file creation and directory management specifically for markdown implementation specs, which is transparently documented in its workflow.
  • [DATA_EXFILTRATION]: The skill reads existing repository code to ensure that the technical design is accurate and consistent with the current codebase; no data is sent externally.
  • [EXTERNAL_DOWNLOADS]: The skill references external sources to verify version information for dependencies, which is a benign informational check.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 19, 2026, 02:44 PM
Security Audit — agent-trust-hub — spec