🐝 Add interactive launch sequence to Phase 0

Replace plain text prompts with a step-by-step interactive launch:
1. Banner displayed immediately
2. ask_user for mission (freeform) if not inline
3. ask_user with 3 pre-selected choices (50/100/250) if not inline
4. Mission briefing + countdown

Inline shortcuts still work — skip prompts when scale/task provided.
Updated both SKILL.md and agent.md. Synced .github copy.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Author: Gregg Cochran

.github/skills/swarm-command/SKILL.md

@@ -36,36 +36,65 @@ Parse the user's input for:
 1. **Scale**: `ss-50`, `ss-100` (default), or `ss-250` — if provided inline
 2. **Task**: Everything after the scale identifier, or the full message if no scale given
 
-If no task provided, ask: "🐝 **Swarm Command ready.** What's the mission?"
+### Interactive Launch Sequence
 
-If no scale provided inline, use ask_user to prompt:
+**Step 1 — Display the launch banner immediately:**
 
 ```
-🐝 Choose your swarm size:
+🐝 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+   S W A R M   C O M M A N D
+   Multi-Model Consensus Orchestrator
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+   50–250 agents · 16 models · one mission
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+```
+
+**Step 2 — If no task provided inline, ask for the mission:**
+
+Use ask_user: "🐝 What's the mission?"
+(Freeform text — the user describes what they want the swarm to do.)
+
+**Step 3 — If no scale provided inline, ask the user to choose swarm size:**
 
-  SS-50   (~52 agents)   ⚡ Fast — single-focus tasks
-  SS-100  (~89 agents)   🎯 Balanced — most tasks (recommended)
-  SS-250  (~316 agents)  🐝 Full swarm — maximum consensus
+Use ask_user with these choices (SS-100 is the recommended default, listed first):
+
+```
+choices:
+  - "🎯 100 agents — balanced, fits most tasks (Recommended)"
+  - "⚡ 50 agents — fast, single-focus tasks"
+  - "🐝 250 agents — full swarm, maximum consensus"
 ```
 
-Display the mission briefing:
+Map the user's selection:
+- "100 agents" → SS-100 (~89 agents, 5 commanders, 8 reviewers)
+- "50 agents" → SS-50 (~52 agents, 3 commanders, 3 reviewers)
+- "250 agents" → SS-250 (~316 agents, 5 commanders, 10 reviewers, 50 squad leads)
+
+**Step 4 — Display the mission briefing and launch:**
 
 ```
 🐝 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
-   S W A R M   C O M M A N D
-   Multi-Model Consensus Orchestrator
-━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
 
 📋 Mission:    <task summary>
 ⚡ Scale:      <SS-50 | SS-100 | SS-250>
 🤖 Agents:     <agent count>
-🧬 Models:     <model count>
+🧬 Models:     16
 💰 Cost cap:   $<ceiling>
 ⏱️  Timeout:    <timeout>s
 
-Deploying swarm in 5... 4... 3... 2... 1...
+   Deploying swarm in 5... 4... 3... 2... 1...
+
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
 ```
 
+### Shortcut: Inline Launch
+
+If the user provides everything inline, skip the interactive prompts and go straight to the mission briefing. Examples:
+- `swarm command ss-250 "Audit the entire codebase for security vulnerabilities"` → skip Steps 2 & 3
+- `swarm command "Refactor auth to JWT"` → skip Step 2, still ask Step 3
+- `swarm command ss-50` → skip Step 3, still ask Step 2
+
 ---
 
 # PHASE 1 — TASK DECOMPOSITION

agents/swarm-command.agent.md

@@ -28,14 +28,49 @@ When the user gives you a task, execute the SwarmSpeed protocol:
 
 Parse for scale (`ss-50`, `ss-100` default, `ss-250`) and task.
 
-Display mission briefing:
+### Interactive Launch Sequence
+
+**Step 1 — Display launch banner immediately:**
+
 ```
 🐝 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
    S W A R M   C O M M A N D
    Multi-Model Consensus Orchestrator
 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+   50–250 agents · 16 models · one mission
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
 ```
 
+**Step 2 — If no task provided inline, ask:**
+Use ask_user: "🐝 What's the mission?" (freeform text)
+
+**Step 3 — If no scale provided inline, ask with choices:**
+```
+choices:
+  - "🎯 100 agents — balanced, fits most tasks (Recommended)"
+  - "⚡ 50 agents — fast, single-focus tasks"
+  - "🐝 250 agents — full swarm, maximum consensus"
+```
+
+**Step 4 — Display mission briefing and launch:**
+```
+🐝 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+📋 Mission:    <task summary>
+⚡ Scale:      <SS-50 | SS-100 | SS-250>
+🤖 Agents:     <agent count>
+🧬 Models:     16
+💰 Cost cap:   $<ceiling>
+⏱️  Timeout:    <timeout>s
+
+   Deploying swarm in 5... 4... 3... 2... 1...
+
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+```
+
+If user provides everything inline (e.g., `swarm command ss-250 "audit security"`), skip prompts and go straight to briefing.
+
 ## Phase 1 — Task Decomposition
 
 Split the task into up to 5 domains:

skills/swarm-command/SKILL.md

@@ -36,36 +36,65 @@ Parse the user's input for:
 1. **Scale**: `ss-50`, `ss-100` (default), or `ss-250` — if provided inline
 2. **Task**: Everything after the scale identifier, or the full message if no scale given
 
-If no task provided, ask: "🐝 **Swarm Command ready.** What's the mission?"
+### Interactive Launch Sequence
 
-If no scale provided inline, use ask_user to prompt:
+**Step 1 — Display the launch banner immediately:**
 
 ```
-🐝 Choose your swarm size:
+🐝 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+   S W A R M   C O M M A N D
+   Multi-Model Consensus Orchestrator
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+   50–250 agents · 16 models · one mission
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+```
+
+**Step 2 — If no task provided inline, ask for the mission:**
+
+Use ask_user: "🐝 What's the mission?"
+(Freeform text — the user describes what they want the swarm to do.)
+
+**Step 3 — If no scale provided inline, ask the user to choose swarm size:**
 
-  SS-50   (~52 agents)   ⚡ Fast — single-focus tasks
-  SS-100  (~89 agents)   🎯 Balanced — most tasks (recommended)
-  SS-250  (~316 agents)  🐝 Full swarm — maximum consensus
+Use ask_user with these choices (SS-100 is the recommended default, listed first):
+
+```
+choices:
+  - "🎯 100 agents — balanced, fits most tasks (Recommended)"
+  - "⚡ 50 agents — fast, single-focus tasks"
+  - "🐝 250 agents — full swarm, maximum consensus"
 ```
 
-Display the mission briefing:
+Map the user's selection:
+- "100 agents" → SS-100 (~89 agents, 5 commanders, 8 reviewers)
+- "50 agents" → SS-50 (~52 agents, 3 commanders, 3 reviewers)
+- "250 agents" → SS-250 (~316 agents, 5 commanders, 10 reviewers, 50 squad leads)
+
+**Step 4 — Display the mission briefing and launch:**
 
 ```
 🐝 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
-   S W A R M   C O M M A N D
-   Multi-Model Consensus Orchestrator
-━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
 
 📋 Mission:    <task summary>
 ⚡ Scale:      <SS-50 | SS-100 | SS-250>
 🤖 Agents:     <agent count>
-🧬 Models:     <model count>
+🧬 Models:     16
 💰 Cost cap:   $<ceiling>
 ⏱️  Timeout:    <timeout>s
 
-Deploying swarm in 5... 4... 3... 2... 1...
+   Deploying swarm in 5... 4... 3... 2... 1...
+
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
 ```
 
+### Shortcut: Inline Launch
+
+If the user provides everything inline, skip the interactive prompts and go straight to the mission briefing. Examples:
+- `swarm command ss-250 "Audit the entire codebase for security vulnerabilities"` → skip Steps 2 & 3
+- `swarm command "Refactor auth to JWT"` → skip Step 2, still ask Step 3
+- `swarm command ss-50` → skip Step 3, still ask Step 2
+
 ---
 
 # PHASE 1 — TASK DECOMPOSITION