[TASK] Phase 1: Core Infrastructure - Auto-Remediation Workflow

Description

Implement the core infrastructure for automated CI/CD failure detection and remediation:

• Create main GitHub Actions workflow (cicd_auto-remediation.yml)
• Integrate Claude Code CLI with cicd-diagnostics skill
• Set up diagnostic cache management
• Configure workflow to trigger on merge queue failures

This is Phase 1 of the Auto-Remediation EPIC (#34656).

Acceptance Criteria

• created with all jobs (detect-failure, diagnose, issue-management, notify)
• Claude Code CLI properly invoked with cicd-diagnostics skill
• Diagnostic cache (GitHub Actions Cache) configured for state management
• Workflow triggers on merge queue completion events
• ANTHROPIC_API_KEY secret configured in repository
• Repository variables set (AUTO_REMEDIATION_ENABLED, AUTO_REMEDIATION_WORKFLOWS, AUTO_CREATE_ISSUES)
• Test run completes successfully with mock failure
• Diagnostic outputs both diagnosis.json and DIAGNOSIS.md

Priority

High

Additional Context

Files to Create:

Dependencies: Requires ANTHROPIC_API_KEY secret to be configured

Related: Parent EPIC #34656

Start with parent EPIC #34656 and map the requirements to the planned cicd_auto-remediation.yml workflow, including detect-failure, diagnose, issue-management, and notify. Verify how Claude Code CLI, the GitHub Actions diagnostic cache, merge queue completion events, repository settings, and mock failure testing should connect; done includes diagnosis.json and DIAGNOSIS.md from a successful test run.