NEW @W-22393672@ Expose ApexGuru engine in CLI with --target-org flag for login

[nikhil-mittal-165]

Summary

• Adds --target-org (-o) flag to the run command for the ApexGuru engine
• Wires target-org through CodeAnalyzerConfigFactory so the apexguru engine receives it via config overrides
• Adds the @salesforce/code-analyzer-apexguru-engine dependency and registers it in EnginePluginsFactory

Test plan

• Unit tests pass locally (npm test)
• Branch coverage above 90% threshold (91.6%)
• Lint clean
• CI green on the new PR

[aruntyagiTutu]

This PR adds the --target-org flag for ApexGuru engine integration. The implementation follows good patterns for config merging but has a critical validation gap per Code Analyzer PR Review guidelines.

✅ Positive aspects:

• Flag properly defined with -o short option and clear documentation
• Target org wired through RunInput → RunAction → ConfigFactory → engine config
• Config factory correctly handles CLI override merging with YAML
• Good test coverage (91.6%, above 90% threshold)
• Tests cover flag parsing and config merging

❌ CRITICAL ISSUE - Validation (#3 in guidelines)

Missing upfront validation:
The PR violates guideline #3: "Token/credential validation upfront. If the workflow requires a GitHub token or org connection, validate it exists and has correct permissions before proceeding."

Current flow:

1. User provides --target-org myorg
2. Flag parsed → RunAction → ConfigFactory → config → ApexGuru engine
3. Org validation only happens deep in engine execution (if at all)

Required change:
Add validation at the boundary (RunAction.execute) BEFORE config creation:

public async execute(input: RunInput): Promise<void> {
    // Validate org authentication upfront
    if (input['target-org']) {
        await this.validateTargetOrg(input['target-org']);
    }
    
    const cliOverrides = // ... existing code
}

private async validateTargetOrg(orgNameOrAlias: string): Promise<void> {
    // Check if org exists and is authenticated
    // Throw with actionable error message if not:
    // - "Org 'myorg' not found. Run 'sf org list' to see available orgs."
    // - "Org 'myorg' not authenticated. Run 'sf org login web --alias myorg'."
}

Test at line 89 explicitly catches and ignores failures with "ApexGuru may fail if org is not authenticated, which is expected" - this confirms validation happens too late.

Secondary issue - Config logic complexity:
Line 66: if (disableSuppressionExplicitlySet && !cliOverrides.targetOrg) - why does targetOrg presence affect suppressions precedence? This coupling is unexpected and makes the code harder to reason about.

Requesting changes to add upfront org validation per guidelines.

[aruntyagiTutu]

LGTM ✅

All validation concerns addressed. The PR now properly validates target-org upfront.

✅ Changes since last review:

• Validation added: validateTargetOrg() now called at the start of RunAction.execute() before expensive config creation
• Uses AuthInfo.create(): Proper org authentication check using Salesforce SDK
• Actionable error messages: Clear guidance on what to do if validation fails
  • Lists available orgs with sf org list
  • Shows how to authenticate with sf org login web
• Excellent test coverage: 5 new tests covering:
  • Happy path (org authenticated)
  • Error path (org not authenticated with clear error message)
  • No validation when flag not provided
  • Validation happens before config creation
  • Config receives targetOrg override

Original positive aspects still apply:

• Flag properly defined with -o short option
• Good documentation in messages/run-command.md
• Config factory correctly merges CLI override with YAML
• ApexGuru engine integration properly wired

This now follows Code Analyzer guideline #3 (validation at boundary).