Dex Improve

📢 Claude Code Updates (since [last check date])

=== NEW CAPABILITIES ===

1. [Feature Name]
   - What it does: [description]
   - How it could help: [specific Dex improvement]
   - Effort to implement: Low/Medium/High

2. [Feature Name]
   ...

=== SUGGESTED IMPROVEMENTS ===

Based on these new capabilities, here's what you could build:

1. [Improvement name]
   - What: [description]
   - Why: [connects to which pillar/job-to-be-done]
   - Uses: [which new capability]

Want to workshop any of these? (Enter number or describe your own idea)

{
  "last_check": "2026-01-22",
  "last_version_seen": "1.0.30",
  "features_seen": [
    "hooks",
    "sub-agents",
    "skills",
    "commands",
    "mcp"
  ],
  "features_noted": [
    {
      "version": "1.0.30",
      "feature": "Async hooks",
      "date_seen": "2026-01-22"
    }
  ]
}

=== CAPABILITY AUDIT ===

✅ Using Well:
- Commands (12 defined)
- Skills (person-lookup)

⚠️ Underutilized:
- Hooks — No hooks configured. Could automate:
  - Session start: Load context automatically
  - Post-edit: Update related files

❌ Not Using:
- Sub-agents — Could parallelize research tasks
- MCP beyond tasks — Calendar, email integrations possible

=== RECOMMENDATIONS ===

Ranked by impact vs. effort:

1. [High impact, low effort]
   - Add SessionStart hook to load daily context
   - Why: Saves 30 seconds every session
   - Connects to: [pillar]

2. [High impact, medium effort]
   ...

# Dex Improvement: [Title]

**Created:** [Date]
**Status:** Planning → In Progress → Complete
**Pillar:** [pillar name if configured]

## Overview

[2-3 sentence summary of what we're improving and why]

## Requirements

- [ ] Requirement 1
- [ ] Requirement 2
- [ ] Requirement 3

## Capability Analysis

| Requirement | Implementation | Feature Type | Rationale |
|-------------|----------------|--------------|-----------|
| [req] | [how] | Hook/Sub-agent/Skill/Command | [why this approach] |

## Recommended Approach

[Which Claude Code features to use and why. Include trade-offs considered.]

## Implementation Steps

### Phase 1: [Foundation]
1. [Step]
2. [Step]

### Phase 2: [Core]
1. [Step]
2. [Step]

## Files to Create/Modify

| File | Action | Purpose |
|------|--------|---------|
| `path/to/file` | Create/Modify | [what] |

## Compound Opportunities

[What else we could solve while building this]

- [ ] [Adjacent improvement 1]
- [ ] [Adjacent improvement 2]

## Acceptance Criteria

- [ ] [Testable outcome 1]
- [ ] [Testable outcome 2]

## Questions Resolved

- Q: [Question that came up]
- A: [How we decided]

/dex-improve "Better meeting action item capture"

> **Understanding:** You want to improve how action items flow from meetings 
> into your task system. This touches workflows (meeting processing) and 
> task management (capture and routing).
>
> **Current state** (from Learnings):
> - Meeting notes go to `00-Inbox/Meetings/`
> - Action items manually extracted during review
>
> **Capability scan:**
> - PostToolUse hook could auto-trigger processing after meeting note creation
> - Sub-agent could parallel process: action items, follow-ups, person updates
>
> **Questions:**
> 1. Where are items getting lost—during capture, or in follow-through?
> 2. Would auto-drafted follow-up emails be valuable?

/dex-improve
> [selects "What's new?"]

> 📢 Claude Code Updates (since 2026-01-15)
>
> === NEW CAPABILITIES ===
> 1. Async Hooks — Hooks can now run without blocking
>    - How it could help: Faster session startup
>    - Effort: Low
>
> === SUGGESTED IMPROVEMENTS ===
> 1. Non-blocking context loading
>    - What: Load learnings/preferences in background
>    - Why: Snappier session starts
>
> Want to workshop any of these?

/dex-improve
> [selects "Full audit"]

> === CAPABILITY AUDIT ===
>
> ✅ Using Well: Commands (12), Skills (1)
> ⚠️ Underutilized: Hooks (0 configured)
> ❌ Not Using: Sub-agents, Advanced MCP
>
> === TOP RECOMMENDATION ===
> Add SessionStart hook — saves context loading time every session
>
> Want me to workshop this improvement?