Add GitHub template governance guidance for AI agents #1078

Describe the task

We need to document issue and PR template enforcement for AI agents to ensure programmatic issue/PR creation follows the organization's governance standards.

The repository has reactive issue template enforcement and preventative PR template enforcement, but AI agents lack clear guidance on:

1. How to determine issue type
2. Which template file to use
3. How to include required sections (Definition of Ready & Done)
4. How to validate before creation

Expected outcome

• AGENTS.md updated with comprehensive GitHub template governance section
• CLAUDE.md updated with quick reference for issue types
• AI agents can create compliant issues/PRs programmatically
• Pre-flight validation prevents post-creation flagging

Additional Context

Analysis shows the enforcement workflow exists but is poorly documented. AI currently bypasses templates when using gh CLI, causing issues to be flagged with status:needs-more-info.

Definition of Ready (DoR)

• Task is clearly described with expected outcome
• Related issue/documentation referenced (issue Branch cleanup automation: fix critical safety issues #1069 analysis)
• Scope is well-defined (2 files: AGENTS.md, CLAUDE.md)
• Ready for development

Definition of Done (DoD)

• AGENTS.md updated with comprehensive guidance
• CLAUDE.md updated with quick reference table
• Code examples provided (JavaScript validation, bash example)
• All 25 issue types documented with template mappings
• Documentation follows project standards (UK English, markdown)
• No regressions in existing documentation
• PR reviewed and approved
• PR merged into develop
• Task issue closed