Migrate from Claude Code to OpenCode structure

- Move legacy content to legacy/ folder (old, old2, docs, learnings, scripts)
- Create new .opencode/ structure with skills/, tools/, agents/ folders
- Update Makefile to symlink to ~/.config/opencode/ instead of ~/.claude/
- Update Makefile to manage skills, tools, and agents (remove settings.json)
- Simplify install/uninstall (no backup logic)
- Add README.md documenting the new structure
- Keep settings.json as historical reference

legacy/old2/skills/templates/user-invocable-skill.md

@@ -0,0 +1,65 @@
+---
+name: skill-name
+description: >
+  Clear description of what this skill does and when to use it.
+  Use when [specific trigger conditions] or when user says /skill-name.
+model: haiku
+argument-hint: <required-param> [optional-param]
+user-invocable: true
+# context: fork  # Use for skills needing isolated context
+# allowed-tools:  # Restrict tools if needed
+#   - Read
+#   - Bash
+---
+
+# Skill Title
+
+@~/.claude/skills/relevant-background-skill/SKILL.md
+
+Brief intro explaining the skill's purpose (1-2 sentences max).
+
+## Process
+
+1. **First step**: Clear action with specific command or instruction
+   - `command or tool to use`
+   - What to look for or validate
+
+2. **Second step**: Next action
+   - Specific details
+   - Expected output
+
+3. **Ask for approval** before significant actions
+   - Show what will be created/modified
+   - Wait for user confirmation
+
+4. **Execute** the approved actions
+   - Run commands/create files
+   - Handle errors gracefully
+
+5. **Present results** with links and summary
+   - Structured output (table or list)
+   - Links to created resources
+
+## Guidelines
+
+- Keep responses concise
+- Use structured output (tables, lists)
+- No preambles or sign-offs
+- Validate inputs before acting
+
+## Output Format
+
+Use this structure for responses:
+
+\`\`\`
+## Summary
+[1-2 sentences]
+
+## Results
+| Item | Status | Link |
+|------|--------|------|
+| ... | ... | ... |
+
+## Next Steps
+- ...
+\`\`\`