create-byan-agent 1.0.1 → 1.0.3

package/templates/_bmad/bmb/workflows/byan/quick-create-workflow.md

@@ -0,0 +1,450 @@
+# BYAN Quick Create Workflow
+
+**Workflow:** Quick Agent Creation (Minimal Questions)  
+**Duration:** 10 minutes  
+**Methodology:** Merise Agile + TDD (with sensible defaults)
+
+---
+
+## OVERVIEW
+
+Quick Create is for experienced users who want to create agents fast with minimal interview.
+
+**Key Differences from Full Interview:**
+- Uses existing ProjectContext (no Phase 1 or 2)
+- Assumes sensible defaults
+- Only asks essential questions
+- 10 min vs 30-45 min
+
+**When to Use:**
+- You've already done a full interview for the project
+- You want to create multiple agents quickly
+- You're comfortable with BMAD agent structure
+- You know exactly what you need
+
+**When NOT to Use:**
+- First agent for a new project → Use full interview
+- Complex or critical agent → Use full interview
+- Uncertain about requirements → Use full interview
+
+---
+
+## PREREQUISITES
+
+**Required:**
+- ProjectContext must exist for this project
+- User must know agent role and capabilities
+
+**Validate before starting:**
+```
+"Quick Create requires an existing project context.
+
+Available projects:
+[List all ProjectContext with project_name]
+
+Which project is this agent for? (or type 'new' for full interview)"
+```
+
+If no ProjectContext exists:
+```
+"No project context found. Quick Create requires project context.
+
+You have two options:
+1. [INT] Start full interview (30-45 min) to create project context
+2. [EXIT] Exit and come back later
+
+Which do you prefer?"
+```
+
+---
+
+## QUICK CREATE FLOW
+
+### Step 1: Load Project Context
+
+```
+"Loading project context for: {project_name}
+
+PROJECT: {project_description}
+DOMAIN: {domain}
+GLOSSARY: {n} concepts
+TECH STACK: {stack}
+
+Ready to create a new agent for this project.
+"
+```
+
+### Step 2: Agent Name & Role
+
+**Q1: Agent Name (kebab-case)**
+```
+"What's the agent name? (kebab-case, e.g., 'backend-expert', 'test-specialist')
+
+Name: _____
+"
+```
+
+Validate:
+- Must be kebab-case: ^[a-z0-9]+(-[a-z0-9]+)*$
+- Must be unique (RG-AGT-001)
+- Suggest if name is too generic:
+  ```
+  "The name 'helper' is very generic. More specific names are better.
+  Suggestions based on your domain:
+  - {domain}-backend-expert
+  - {domain}-api-specialist
+  - {domain}-test-engineer
+  
+  Use 'helper' anyway? (yes/no)"
+  ```
+
+**Q2: Role (one sentence)**
+```
+"What's this agent's role? (one sentence, like a job title)
+
+Example: 'Backend API specialist who ensures RESTful best practices'
+
+Role: _____
+"
+```
+
+### Step 3: Capabilities (Minimum 3)
+
+```
+"What are the 3 most critical capabilities? (minimum 3, can add more)
+
+Think in terms of CREATE, ANALYZE, REVIEW, OPTIMIZE, TEACH.
+
+Examples:
+- 'Generate API endpoints with OpenAPI specs'
+- 'Review database queries for N+1 problems'
+- 'Create integration tests with realistic fixtures'
+
+Capability 1: _____
+Capability 2: _____
+Capability 3: _____
+More? (press Enter to skip)
+"
+```
+
+Validate RG-AGT-002: >= 3 capabilities
+
+**Auto-categorize capabilities:**
+```python
+capability_categories = {
+    "generate|create|build|scaffold": "create",
+    "analyze|review|audit|check": "analyze",
+    "optimize|improve|refactor": "optimize",
+    "teach|explain|guide": "teach"
+}
+```
+
+### Step 4: Mantras (Auto-suggest + Custom)
+
+```
+"Which mantras should this agent prioritize?
+
+I'll suggest 5 based on your role and capabilities. 
+You can accept these or customize.
+
+SUGGESTED MANTRAS:
+[Auto-suggest based on role/capabilities]
+1. Mantra #{id}: {name} - {why relevant}
+2. Mantra #{id}: {name} - {why relevant}
+3. Mantra #{id}: {name} - {why relevant}
+4. Mantra #{id}: {name} - {why relevant}
+5. Mantra #{id}: {name} - {why relevant}
+
+Accept these suggestions? (yes/no/show-all)
+"
+```
+
+**Auto-suggestion logic:**
+```python
+def suggest_mantras(role, capabilities):
+    suggestions = []
+    
+    # Always include
+    suggestions.append(37)  # Ockham's Razor
+    suggestions.append(39)  # Consequences
+    
+    # Role-based
+    if "backend" in role.lower():
+        suggestions.append(4)   # Fail Fast
+        suggestions.append(20)  # Performance
+        suggestions.append(18)  # TDD
+    elif "test" in role.lower():
+        suggestions.append(18)  # TDD
+        suggestions.append(19)  # Test Behavior
+        suggestions.append(16)  # Challenge Before Confirm (IA)
+    elif "frontend" in role.lower():
+        suggestions.append(12)  # UX is Priority
+        suggestions.append(20)  # Performance
+        suggestions.append(7)   # KISS
+    
+    # Capability-based
+    for cap in capabilities:
+        if "security" in cap.lower():
+            suggestions.append(21)  # Security by Design
+        if "optimize" in cap.lower():
+            suggestions.append(20)  # Performance
+        if "review" in cap.lower():
+            suggestions.append(16)  # Challenge Before Confirm (IA)
+    
+    # Dedupe and take first 5
+    return list(set(suggestions))[:5]
+```
+
+If user says "show-all":
+```
+"Full list of 64 mantras available in {project-root}/_bmad-output/guide-reference-rapide-merise-agile-tdd.md
+
+Or I can list them here by category:
+1. Show Conception Mantras (39)
+2. Show AI Agent Mantras (25)
+3. Keep suggestions
+
+What would you like?"
+```
+
+### Step 5: Communication Style (Quick)
+
+```
+"Communication style for this agent? Pick one:
+
+1. CONCISE - Direct, minimal words, assumes expertise
+2. EDUCATIONAL - Detailed explanations, examples, teaches as it works
+3. BALANCED - Mix of both, adapts to context
+
+Your choice (1/2/3): ___
+"
+```
+
+### Step 6: Quick Validation
+
+```
+"Let me confirm what we're creating:
+
+AGENT: {agent_name}
+ROLE: {role}
+CAPABILITIES: {n} capabilities
+  - {capability_1}
+  - {capability_2}
+  - {capability_3}
+  ...
+MANTRAS: {n} mantras
+  - Mantra #{id}: {name}
+  ...
+STYLE: {communication_style}
+PROJECT: {project_name}
+
+This will take ~30 seconds to generate.
+
+Proceed? (yes/no/edit)"
+```
+
+If "edit":
+```
+"What needs adjustment?
+1. Agent name
+2. Role
+3. Capabilities
+4. Mantras
+5. Communication style
+6. Cancel
+
+Choice: ___"
+```
+
+---
+
+## GENERATION
+
+### Create AgentSpec
+
+```python
+agent_spec = AgentSpec.create(
+    context_id=context.id,
+    agent_name=answers['agent_name'],
+    role=answers['role'],
+    responsibilities=[],  # Auto-derive from capabilities
+    knowledge_business=context.glossaire.keys(),  # Inherit from context
+    knowledge_technical=context.stack_tech,  # Inherit from context
+    capabilities=answers['capabilities'],
+    mantras_applied=answers['mantras'],
+    communication_style=answers['communication_style'],
+    use_cases=[],  # Will be empty for quick create
+    status="draft"
+)
+
+# Validate
+agent_spec.validate()  # RG-AGT-001, 002, 003 checked
+
+# Save
+agent_spec.save()
+```
+
+**Handle validation failures:**
+```
+"Validation failed:
+- {error_1}
+- {error_2}
+
+Quick fixes:
+{suggested_fixes}
+
+Try again? (yes/no)"
+```
+
+---
+
+## COMPLETION
+
+### Success Message
+
+```
+"Agent created successfully!
+
+AGENT: {agent_name}
+STATUS: Validated and ready for file generation
+LOCATION: _bmad/{module}/agents/{agent_name}.md (pending generation)
+
+NEXT STEPS:
+1. Generate agent file for platform
+2. Test the agent
+3. Iterate based on usage
+
+Generate files now for which platform(s)?
+- copilot (GitHub Copilot CLI)
+- vscode (VSCode extension)
+- claude (Claude Code)
+- codex (Codex)
+- all (all platforms)
+- later (exit, generate later)
+
+Platform: ____
+"
+```
+
+### If "later":
+
+```
+"Agent spec saved! You can generate files anytime with:
+
+[BYAN Menu] > Generate Files
+
+Or via CLI:
+  byan generate {agent_name} --platform=copilot
+
+Session summary saved to:
+{output_folder}/quick-create-{agent_name}-{timestamp}.md
+"
+```
+
+---
+
+## DEFAULTS APPLIED
+
+Quick Create uses these defaults:
+
+**Responsibilities:** Auto-derived from capabilities
+```python
+def derive_responsibilities(capabilities):
+    return [cap['description'] for cap in capabilities]
+```
+
+**Use Cases:** Left empty (can be added later)
+
+**Knowledge:** Inherited from ProjectContext
+- Business: All glossary concepts
+- Technical: Full tech stack
+
+**Menu Structure:** Standard BMAD menu
+```
+[MH] Menu Help
+[CH] Chat
+[TASK] Execute primary task
+[EXIT] Dismiss
+```
+
+**Persona Template:** Based on role + communication style
+
+**Anti-patterns:** Standard set from 64 mantras
+
+---
+
+## ERROR HANDLING
+
+**No ProjectContext:**
+```
+"ERROR: No project context found.
+
+Quick Create requires existing context. Please:
+1. Run full interview first: [INT]
+2. Or import existing context
+
+Cannot proceed with Quick Create."
+```
+
+**Duplicate Agent Name:**
+```
+"ERROR: Agent '{agent_name}' already exists.
+
+Existing agents in this project:
+- {agent_1}
+- {agent_2}
+...
+
+Choose a different name or edit existing agent."
+```
+
+**Insufficient Capabilities:**
+```
+"ERROR: Need at least 3 capabilities (you provided {n}).
+
+Please provide {3-n} more capabilities."
+```
+
+---
+
+## COMPARISON: Full Interview vs Quick Create
+
+| Aspect | Full Interview | Quick Create |
+|--------|----------------|--------------|
+| Duration | 30-45 min | 10 min |
+| Project Context | Created | Reused |
+| Business Glossary | Interactive creation | Inherited |
+| Capabilities | Deep exploration | Quick list |
+| Mantras | Justified selection | Auto-suggested |
+| Use Cases | Minimum 3 required | Optional |
+| Validation | Extensive | Basic |
+| Suitable for | First agent, critical agents | Additional agents |
+
+---
+
+## SUCCESS CRITERIA
+
+✅ ProjectContext exists
+✅ Agent name unique and valid format
+✅ RG-AGT-002: >= 3 capabilities
+✅ RG-AGT-003: >= 5 mantras
+✅ AgentSpec validated
+✅ User confirmed creation
+✅ Completed in < 15 minutes
+
+---
+
+## NEXT WORKFLOW
+
+After Quick Create, automatically offer:
+```
+"Agent created! 
+
+Want to enhance it?
+- [VA] Validate against all 64 mantras
+- [EA] Edit to add use cases or refine
+- Generate files now
+- Exit
+
+What would you like to do?"
+```

package/templates/_bmad/bmb/workflows/byan/templates/base-agent-template.md

@@ -0,0 +1,79 @@
+---
+name: "{agent_name}"
+description: "{agent_description}"
+---
+
+You must fully embody this agent's persona and follow all activation instructions exactly as specified. NEVER break character until given an exit command.
+
+```xml
+<agent id="{agent_id}" name="{agent_display_name}" title="{agent_title}" icon="{agent_icon}">
+<activation critical="MANDATORY">
+      <step n="1">Load persona from this current agent file (already in context)</step>
+      <step n="2">🚨 IMMEDIATE ACTION REQUIRED - BEFORE ANY OUTPUT:
+          - Load and read {project-root}/_bmad/{module}/config.yaml NOW
+          - Store ALL fields as session variables: {user_name}, {communication_language}, {output_folder}
+          - VERIFY: If config not loaded, STOP and report error to user
+          - DO NOT PROCEED to step 3 until config is successfully loaded and variables stored
+      </step>
+      <step n="3">Remember: user's name is {user_name}</step>
+      
+      <step n="4">Show greeting using {user_name} from config, communicate in {communication_language}, then display numbered list of ALL menu items from menu section</step>
+      <step n="5">Let {user_name} know they can type command `/bmad-help` at any time to get advice on what to do next, and that they can combine that with what they need help with <example>`/bmad-help {example_help_query}`</example></step>
+      <step n="6">STOP and WAIT for user input - do NOT execute menu items automatically - accept number or cmd trigger or fuzzy command match</step>
+      <step n="7">On user input: Number → process menu item[n] | Text → case-insensitive substring match | Multiple matches → ask user to clarify | No match → show "Not recognized"</step>
+      <step n="8">When processing a menu item: Check menu-handlers section below - extract any attributes from the selected menu item (workflow, exec, tmpl, data, action, validate-workflow) and follow the corresponding handler instructions</step>
+
+      <menu-handlers>
+              <handlers>
+          <handler type="exec">
+        When menu item or handler has: exec="path/to/file.md":
+        1. Read fully and follow the file at that path
+        2. Process the complete file and follow all instructions within it
+        3. If there is data="some/path/data-foo.md" with the same item, pass that data path to the executed file as context.
+      </handler>
+        </handlers>
+      </menu-handlers>
+
+    <rules>
+      <r>ALWAYS communicate in {communication_language} UNLESS contradicted by communication_style.</r>
+      <r>Stay in character until exit selected</r>
+      <r>Display Menu items as the item dictates and in the order given.</r>
+      <r>Load files ONLY when executing a user chosen workflow or a command requires it, EXCEPTION: agent activation step 2 config.yaml</r>
+      {custom_rules}
+    </rules>
+</activation>
+
+<persona>
+    <role>{role}</role>
+    <identity>{identity}</identity>
+    <communication_style>{communication_style}</communication_style>
+    <principles>{principles}</principles>
+    {mantras_section}
+  </persona>
+  
+  {knowledge_base_section}
+  
+  <menu>
+    <item cmd="MH or fuzzy match on menu or help">[MH] Redisplay Menu Help</item>
+    <item cmd="CH or fuzzy match on chat">[CH] Chat with {agent_display_name} about anything</item>
+    {custom_menu_items}
+    <item cmd="EXIT or fuzzy match on exit, leave, goodbye or dismiss agent">[EXIT] Dismiss {agent_display_name}</item>
+  </menu>
+  
+  <capabilities>
+    {capabilities_list}
+  </capabilities>
+  
+  <anti_patterns>
+    {anti_patterns_list}
+  </anti_patterns>
+  
+  <exit_protocol>
+    When user selects EXIT:
+    1. {exit_step_1}
+    2. {exit_step_2}
+    3. {exit_step_3}
+    4. Return control to user
+  </exit_protocol>
+</agent>
+```