Release v5.0.0 - Agent Coordination & Dashboard

.claude/commands/agent.md

@@ -97,14 +97,31 @@ Assemble multiple agents for collaborative work.
 3. Set up handoff points
 4. Configure aggregation of results
 
-**Team composition example:**
+**Team composition examples:**
 
 ```yaml
+# Comprehensive code review
 team: code-review-team
 agents:
-  - code-reviewer     # Quality check
-  - security-reviewer # Security audit
-  - architect        # Architecture review
+  - code-reviewer       # Quality check
+  - security-reviewer   # Security audit
+  - quality-validator   # Metrics analysis
+workflow: sequential
+
+# Debug and fix
+team: debug-fix-team
+agents:
+  - debugger            # Find root cause
+  - test-writer         # Write regression test
+  - refactorer          # Clean up fix
+workflow: sequential
+
+# Feature planning
+team: planning-team
+agents:
+  - planner             # Requirements discovery
+  - architect           # Technical design
+  - performance-analyst # Scalability review
 workflow: sequential
 ```
 
@@ -127,12 +144,20 @@ View all available agents and their capabilities.
 
 **Available agents:**
 
-| Agent                  | Role            | Use Case                |
-| ---------------------- | --------------- | ----------------------- |
-| **architect**          | System design   | Architecture decisions  |
-| **code-reviewer**      | Quality review  | Code review tasks       |
-| **security-reviewer**  | Security audit  | Security-sensitive code |
-| **empathy-specialist** | Level 4+ work   | Complex user needs      |
+All agents use **Socratic questioning** to guide discovery rather than just providing answers.
+
+| Agent                    | Role               | Use Case                              |
+| ------------------------ | ------------------ | ------------------------------------- |
+| **architect**            | System design      | Architecture decisions, tech choices  |
+| **code-reviewer**        | Quality review     | Code review, design patterns          |
+| **debugger**             | Bug investigation  | Root cause analysis, hypothesis tests |
+| **empathy-specialist**   | Level 4-5 empathy  | Complex user needs, trust building    |
+| **performance-analyst**  | Optimization       | Bottlenecks, profiling, memory issues |
+| **planner**              | Requirements       | Sprint planning, scope discovery      |
+| **quality-validator**    | Code quality       | Complexity, naming, documentation     |
+| **refactorer**           | Code improvement   | Clean code, pattern application       |
+| **security-reviewer**    | Security audit     | Vulnerabilities, attack scenarios     |
+| **test-writer**          | Test design        | Edge cases, test strategy, TDD        |
 
 ---
 
@@ -154,17 +179,33 @@ Run a specific agent for a task.
 
 ---
 
-## When NOT to Use This Hub
+## Cross-Hub Agent Integration
+
+Agents can also be invoked from relevant hubs:
+
+| Hub          | Related Agents                                  |
+| ------------ | ----------------------------------------------- |
+| `/dev`       | debugger, code-reviewer, refactorer             |
+| `/testing`   | test-writer, quality-validator                  |
+| `/workflows` | (runs automated multi-stage workflows)          |
+| `/plan`      | planner, architect                              |
+| `/docs`      | code-reviewer (documentation review)            |
+| `/release`   | security-reviewer, quality-validator            |
+
+## When to Use This Hub Directly
+
+Use `/agent` when you want to:
 
-| If you need...        | Use instead |
-| --------------------- | ----------- |
-| Debug code            | `/dev`      |
-| Run tests             | `/testing`  |
-| Plan implementation   | `/workflow` |
-| Manage context/memory | `/context`  |
+- Create a new custom agent
+- Compose an agent team for collaborative work
+- Invoke a specific agent by name
+- See all available agents and their capabilities
 
 ## Related Hubs
 
-- `/workflow` - Development workflows
+- `/workflows` - Run automated AI workflows (security-audit, bug-predict, etc.)
+- `/plan` - Development planning (uses planner, architect)
+- `/dev` - Development tools (uses debugger, code-reviewer)
+- `/testing` - Testing workflows (uses test-writer)
 - `/context` - Context and memory management
 - `/learning` - Pattern learning

.claude/commands/context.md

@@ -5,6 +5,19 @@ category: hub
 aliases: [ctx]
 tags: [context, memory, state, profile, hub]
 version: "2.0"
+question:
+  header: "Context"
+  question: "What context operation do you need?"
+  multiSelect: false
+  options:
+    - label: "📊 Show status"
+      description: "Display current session state, tasks, and context"
+    - label: "💭 View memory"
+      description: "Browse and search learned patterns and preferences"
+    - label: "💾 Save state"
+      description: "Preserve current state for later resumption"
+    - label: "📝 Edit CLAUDE.md"
+      description: "Update project memory and instructions"
 ---
 
 # Context Management
@@ -139,7 +152,8 @@ View session status or update your preferences.
 | ------------------ | ------------ |
 | Create commits     | `/dev`       |
 | Run tests          | `/testing`   |
-| Plan features      | `/workflow`  |
+| Plan features      | `/plan`      |
+| Run workflows      | `/workflows` |
 | Learn patterns     | `/learning`  |
 
 ## Related Hubs