Platform Apex Test Run

This handles Apex test execution and the full test-fix cycle: running tests, parsing coverage, diagnosing failures, and looping until things pass. It knows when to run a single class versus local tests, enforces disciplined patterns like SeeAllData=false and 251-record bulk tests, and reports results in a structured format with root cause and next steps. The skill includes templates for mocks, data factories, and a Python script that parses test JSON for automated fix loops. It deliberately stays in its lane, handing off production code changes to platform-apex-generate and agent testing to agentforce-test. The 120-point scoring system gives you a quick read on test quality. Useful when coverage drops after a refactor or CI starts failing on tests that passed locally.

Install to Claude Code

npx -y skills add forcedotcom/sf-skills --skill platform-apex-test-run --agent claude-code