sf-testing

Pass

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill operates by executing standard Salesforce CLI commands (sf apex run test) against the user's authenticated environment.
  • [SAFE]: The Python hook script (parse-test-results.py) is included within the skill package. It analyzes the tool's output locally to provide structured feedback to the agent, helping identify error types like NullPointerException or DmlException without any external network calls.
  • [SAFE]: All included Apex templates and mocking patterns (DML mocking, StubProvider examples) follow established Salesforce community best practices and are designed for performance and isolation.
  • [SAFE]: The 'agentic fix loop' is a high-level instructional workflow for the agent to interpret results and suggest corrections to the code. It does not involve autonomous execution of untrusted remote code.
  • [SAFE]: Metadata and documentation are transparent and correctly attribute community sources (e.g., James Simone, Beyond the Cloud).
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 14, 2026, 01:51 AM