feat: agentskills build output

[teddyaryono]

Summary

Add AI-DLC rules converted to AgentSkills format, enabling the AI-DLC workflow to be used with AgentSkills-compatible agents.

Changes

• Add build/skills/aidlc/ directory containing 30 files with the full AI-DLC ruleset converted to AgentSkills format, covering:
  • SKILL.md — main skill entry point
  • common/ — shared rules (error handling, terminology, session continuity, etc.)
  • inception/ — planning phase rules (requirements analysis, user stories, workflow planning, etc.)
  • construction/ — implementation phase rules (code generation, functional design, build & test, etc.)
  • extensions/ — opt-in extensions (security baseline, property-based testing)
  • operations/ — deployment phase rules
• Add "Build Output" section to README documenting the build/ directory, its purpose, and a note that files should not be edited directly

User experience

Users can consume the AI-DLC workflow through the AgentSkills standard by copying the build/skills/aidlc/ directory into their agent's skills configuration directory. For example:

• Claude Code: Copy to .claude/skills/
• Kiro: Copy to .kiro/skills/
• Other AgentSkills-compatible agents: Copy to the agent's designated skills directory

Checklist

If your change doesn't seem to apply, please leave them unchecked.

• I have reviewed the contributing guidelines
• I have performed a self-review of this change
• Changes have been tested
• Changes are documented

Test Plan

• Verify build/skills/aidlc/SKILL.md is a valid AgentSkills entry point and references all sub-rules correctly
• Verify the skill file structure matches the AgentSkills format (SKILL.md at root, categorized subdirectories)

Acknowledgment

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of the project license.

[harmjeff]

Thanks for the contribution! The idea of making AI-DLC available in AgentSkills format is great, but we need to fix the approach before merging.

Issue: Checked-in build output creates copy-paste maintenance problems.

Committing 30 generated files directly to the repo means every time the source rules change, someone has to manually regenerate and update these files — or they drift out of sync. This doesn't scale.

What we need instead:

A build tool/script (e.g., scripts/build-agentskills.sh or a CI workflow) that generates the AgentSkills output from the source rules at release time. This way:

1. Source rules remain the single source of truth
2. The build output is always in sync with the latest release
3. No manual copy-paste maintenance burden
4. Contributors only edit source rules, not generated artifacts

The build/ directory should be in .gitignore, and the generation should happen as part of the release pipeline (or on-demand via a script users can run locally).

Please rework this PR to include the conversion tool/script rather than the static output files. Happy to discuss the approach if helpful.