Brian Madison 054b031c1d feat: Complete BMAD agent creation system with install tooling, references, and field guidance
## Overview
This commit represents a complete overhaul of the BMAD agent creation system, establishing clear standards for agent development, installation workflows, and persona design. The changes span documentation, tooling, reference implementations, and field-specific guidance.

## Key Components

### 1. Agent Installation Infrastructure
**New CLI Command: `agent-install`**
- Interactive agent installation with persona customization
- Supports Simple (single YAML), Expert (sidecar files), and Module agents
- Template variable processing with Handlebars-style syntax
- Automatic compilation from YAML to XML (.md) format
- Manifest tracking and IDE integration (Claude Code, Cursor, Windsurf, etc.)
- Source preservation in `_cfg/custom/agents/` for reinstallation

**Files Created:**
- `tools/cli/commands/agent-install.js` - Main CLI command
- `tools/cli/lib/agent/compiler.js` - YAML to XML compilation engine
- `tools/cli/lib/agent/installer.js` - Installation orchestration
- `tools/cli/lib/agent/template-engine.js` - Handlebars template processing

**Compiler Features:**
- Auto-injects frontmatter, activation, handlers, help/exit menu items
- Smart handler inclusion (only includes action/workflow/exec/tmpl handlers actually used)
- Proper XML escaping and formatting
- Persona name customization (e.g., "Fred the Commit Poet")

### 2. Documentation Overhaul
**Deleted Bloated/Outdated Docs (2,651 lines removed):**
- Old verbose architecture docs
- Redundant pattern files
- Outdated workflow guides

**Created Focused, Type-Specific Docs:**
- `src/modules/bmb/docs/understanding-agent-types.md` - Architecture vs capability distinction
- `src/modules/bmb/docs/simple-agent-architecture.md` - Self-contained agents
- `src/modules/bmb/docs/expert-agent-architecture.md` - Agents with sidecar files
- `src/modules/bmb/docs/module-agent-architecture.md` - Workflow-integrated agents
- `src/modules/bmb/docs/agent-compilation.md` - YAML → XML process
- `src/modules/bmb/docs/agent-menu-patterns.md` - Menu design patterns
- `src/modules/bmb/docs/index.md` - Documentation hub

**Net Result:** ~1,930 line reduction while adding MORE value through focused content

### 3. Create-Agent Workflow Enhancements
**Critical Persona Field Guidance Added to Step 4:**
Explains how the LLM interprets each persona field when the agent activates:

- **role** → "What knowledge, skills, and capabilities do I possess?"
- **identity** → "What background, experience, and context shape my responses?"
- **communication_style** → "What verbal patterns, word choice, quirks, and phrasing do I use?"
- **principles** → "What beliefs and operating philosophy drive my choices?"

**Key Insight:** `communication_style` should ONLY describe HOW the agent talks, not restate role/identity/principles. The `communication-presets.csv` provides 60 pure communication styles with NO role/identity/principles mixed in.

**Files Updated:**
- `src/modules/bmb/workflows/create-agent/instructions.md` - Added persona field interpretation guide
- `src/modules/bmb/workflows/create-agent/brainstorm-context.md` - Refined to 137 lines
- `src/modules/bmb/workflows/create-agent/communication-presets.csv` - 60 styles across 13 categories

### 4. Reference Agent Cleanup
**Removed install_config Personality Bloat:**
Understanding: Future installer will handle personality customization, so stripped all personality toggles from reference agents.

**commit-poet.agent.yaml** (Simple Agent):
- BEFORE: 36 personality combinations (3 enthusiasm × 3 depths × 4 styles) = decision fatigue
- AFTER: Single concise persona with pure communication style
- Changed from verbose conditionals to: "Poetic drama and flair with every turn of a phrase. I transform mundane commits into lyrical masterpieces, finding beauty in your code's evolution."
- Reduction: 248 lines → 153 lines (38% reduction)

**journal-keeper.agent.yaml** (Expert Agent):
- Stripped install_config, simplified communication_style
- Shows proper Expert agent structure with sidecar files

**security-engineer.agent.yaml & trend-analyst.agent.yaml** (Module Agents):
- Added header comments explaining WHY Module Agent (design intent, not just location)
- Clarified: Module agents are designed FOR ecosystem integration, not capability-limited

**Files Updated:**
- `src/modules/bmb/reference/agents/simple-examples/commit-poet.agent.yaml`
- `src/modules/bmb/reference/agents/expert-examples/journal-keeper/journal-keeper.agent.yaml`
- `src/modules/bmb/reference/agents/module-examples/security-engineer.agent.yaml`
- `src/modules/bmb/reference/agents/module-examples/trend-analyst.agent.yaml`

### 5. BMM Agent Voice Enhancement
**Gave all 9 BMM agents distinct, memorable communication voices:**

**Mary (analyst)** - The favorite! Changed from generic "systematic and probing" to:
"Treats analysis like a treasure hunt - excited by every clue, thrilled when patterns emerge. Asks questions that spark 'aha!' moments while structuring insights with precision."

**Other Notable Voices:**
- **John (pm):** "Asks 'WHY?' relentlessly like a detective on a case. Direct and data-sharp, cuts through fluff to what actually matters."
- **Winston (architect):** "Speaks in calm, pragmatic tones, balancing 'what could be' with 'what should be.' Champions boring technology that actually works."
- **Amelia (dev):** "Ultra-succinct. Speaks in file paths and AC IDs - every statement citable. No fluff, all precision."
- **Bob (sm):** "Crisp and checklist-driven. Every word has a purpose, every requirement crystal clear. Zero tolerance for ambiguity."
- **Sally (ux-designer):** "Paints pictures with words, telling user stories that make you FEEL the problem. Empathetic advocate with creative storytelling flair."

**Pattern Applied:** Moved behaviors from communication_style to principles, keeping communication_style as PURE verbal patterns.

**Files Updated:**
- `src/modules/bmm/agents/analyst.agent.yaml`
- `src/modules/bmm/agents/pm.agent.yaml`
- `src/modules/bmm/agents/architect.agent.yaml`
- `src/modules/bmm/agents/dev.agent.yaml`
- `src/modules/bmm/agents/sm.agent.yaml`
- `src/modules/bmm/agents/tea.agent.yaml`
- `src/modules/bmm/agents/tech-writer.agent.yaml`
- `src/modules/bmm/agents/ux-designer.agent.yaml`
- `src/modules/bmm/agents/frame-expert.agent.yaml`

### 6. Linting Fixes
**ESLint Compliance:**
- Replaced all `'utf-8'` with `'utf8'` (unicorn/text-encoding-identifier-case)
- Changed `variables.hasOwnProperty(varName)` to `Object.hasOwn(variables, varName)` (unicorn/prefer-object-has-own)
- Replaced `JSON.parse(JSON.stringify(...))` with `structuredClone(...)` (unicorn/prefer-structured-clone)
- Fixed empty YAML mapping values in sample files

**Files Fixed:**
- 7 JavaScript files across agent tooling (compiler, installer, commands, IDE integration)
- 1 YAML sample file

## Architecture Decisions

### Agent Types Are About Architecture, Not Capability
- **Simple:** Self-contained in single YAML (NOT limited in capability)
- **Expert:** Includes sidecar files (templates, docs, etc.)
- **Module:** Designed for BMAD ecosystem integration (workflows, cross-agent coordination)

### Persona Field Separation Critical for LLM Interpretation
The LLM needs distinct fields to understand its role:
- Mixing role/identity/principles into communication_style confuses the persona
- Pure communication styles (from communication-presets.csv) have ZERO role/identity/principles content
- Example DON'T: "Experienced analyst who uses systematic approaches..." (mixing identity + style)
- Example DO: "Systematic and probing. Structures findings hierarchically." (pure style)

### Install-Time vs Runtime Configuration
- Template variables ({{var}}) resolve at compile-time
- Runtime variables ({user_name}, {bmad_folder}) resolve when agent activates
- Future installer will handle personality customization, so agents should ship with single default persona

## Testing
- All linting passes (ESLint with max-warnings=0)
- Agent compilation tested with commit-poet, journal-keeper examples
- Install workflow validated with Simple and Expert agent types
- Manifest tracking and IDE integration verified

## Impact
This establishes BMAD as having a complete, production-ready agent creation and installation system with:
- Clear documentation for all agent types
- Automated compilation and installation
- Strong persona design guidance
- Reference implementations showing best practices
- Distinct, memorable agent voices throughout BMM module

Co-Authored-By: BMad Builder <builder@bmad.dev>
Co-Authored-By: Mary the Analyst <analyst@bmad.dev>
Co-Authored-By: Paige the Tech Writer <tech-writer@bmad.dev>
2025-11-18 21:55:47 -06:00

515 lines
18 KiB
JavaScript

const path = require('node:path');
const { BaseIdeSetup } = require('./_base-ide');
const chalk = require('chalk');
const { getProjectRoot, getSourcePath, getModulePath } = require('../../../lib/project-root');
const { WorkflowCommandGenerator } = require('./shared/workflow-command-generator');
const { TaskToolCommandGenerator } = require('./shared/task-tool-command-generator');
const { AgentCommandGenerator } = require('./shared/agent-command-generator');
const {
loadModuleInjectionConfig,
shouldApplyInjection,
filterAgentInstructions,
resolveSubagentFiles,
} = require('./shared/module-injections');
const { getAgentsFromBmad, getAgentsFromDir } = require('./shared/bmad-artifacts');
/**
* Claude Code IDE setup handler
*/
class ClaudeCodeSetup extends BaseIdeSetup {
constructor() {
super('claude-code', 'Claude Code', true); // preferred IDE
this.configDir = '.claude';
this.commandsDir = 'commands';
this.agentsDir = 'agents';
}
/**
* Collect configuration choices before installation
* @param {Object} options - Configuration options
* @returns {Object} Collected configuration
*/
async collectConfiguration(options = {}) {
const config = {
subagentChoices: null,
installLocation: null,
};
const sourceModulesPath = getSourcePath('modules');
const modules = options.selectedModules || [];
for (const moduleName of modules) {
// Check for Claude Code sub-module injection config in SOURCE directory
const injectionConfigPath = path.join(sourceModulesPath, moduleName, 'sub-modules', 'claude-code', 'injections.yaml');
if (await this.exists(injectionConfigPath)) {
const fs = require('fs-extra');
const yaml = require('js-yaml');
try {
// Load injection configuration
const configContent = await fs.readFile(injectionConfigPath, 'utf8');
const injectionConfig = yaml.load(configContent);
// Ask about subagents if they exist and we haven't asked yet
if (injectionConfig.subagents && !config.subagentChoices) {
config.subagentChoices = await this.promptSubagentInstallation(injectionConfig.subagents);
if (config.subagentChoices.install !== 'none') {
// Ask for installation location
const inquirer = require('inquirer');
const locationAnswer = await inquirer.prompt([
{
type: 'list',
name: 'location',
message: 'Where would you like to install Claude Code subagents?',
choices: [
{ name: 'Project level (.claude/agents/)', value: 'project' },
{ name: 'User level (~/.claude/agents/)', value: 'user' },
],
default: 'project',
},
]);
config.installLocation = locationAnswer.location;
}
}
} catch (error) {
console.log(chalk.yellow(` Warning: Failed to process ${moduleName} features: ${error.message}`));
}
}
}
return config;
}
/**
* Cleanup old BMAD installation before reinstalling
* @param {string} projectDir - Project directory
*/
async cleanup(projectDir) {
const fs = require('fs-extra');
const bmadCommandsDir = path.join(projectDir, this.configDir, this.commandsDir, 'bmad');
if (await fs.pathExists(bmadCommandsDir)) {
await fs.remove(bmadCommandsDir);
console.log(chalk.dim(` Removed old BMAD commands from ${this.name}`));
}
}
/**
* Setup Claude Code IDE configuration
* @param {string} projectDir - Project directory
* @param {string} bmadDir - BMAD installation directory
* @param {Object} options - Setup options
*/
async setup(projectDir, bmadDir, options = {}) {
// Store project directory for use in processContent
this.projectDir = projectDir;
console.log(chalk.cyan(`Setting up ${this.name}...`));
// Clean up old BMAD installation first
await this.cleanup(projectDir);
// Create .claude/commands directory structure
const claudeDir = path.join(projectDir, this.configDir);
const commandsDir = path.join(claudeDir, this.commandsDir);
const bmadCommandsDir = path.join(commandsDir, 'bmad');
await this.ensureDir(bmadCommandsDir);
// Generate agent launchers using AgentCommandGenerator
// This creates small launcher files that reference the actual agents in .bmad/
const agentGen = new AgentCommandGenerator(this.bmadFolderName);
const { artifacts: agentArtifacts, counts: agentCounts } = await agentGen.collectAgentArtifacts(bmadDir, options.selectedModules || []);
// Create directories for each module
const modules = new Set();
for (const artifact of agentArtifacts) {
modules.add(artifact.module);
}
for (const module of modules) {
await this.ensureDir(path.join(bmadCommandsDir, module));
await this.ensureDir(path.join(bmadCommandsDir, module, 'agents'));
}
// Write agent launcher files
const agentCount = await agentGen.writeAgentLaunchers(bmadCommandsDir, agentArtifacts);
// Process Claude Code specific injections for installed modules
// Use pre-collected configuration if available, or skip if already configured
if (options.preCollectedConfig && options.preCollectedConfig._alreadyConfigured) {
// IDE is already configured from previous installation, skip prompting
// Just process with default/existing configuration
await this.processModuleInjectionsWithConfig(projectDir, bmadDir, options, {});
} else if (options.preCollectedConfig) {
await this.processModuleInjectionsWithConfig(projectDir, bmadDir, options, options.preCollectedConfig);
} else {
await this.processModuleInjections(projectDir, bmadDir, options);
}
// Skip CLAUDE.md creation - let user manage their own CLAUDE.md file
// await this.createClaudeConfig(projectDir, modules);
// Generate workflow commands from manifest (if it exists)
const workflowGen = new WorkflowCommandGenerator(this.bmadFolderName);
const { artifacts: workflowArtifacts } = await workflowGen.collectWorkflowArtifacts(bmadDir);
// Write only workflow-command artifacts, skip workflow-launcher READMEs
let workflowCommandCount = 0;
for (const artifact of workflowArtifacts) {
if (artifact.type === 'workflow-command') {
const moduleWorkflowsDir = path.join(bmadCommandsDir, artifact.module, 'workflows');
await this.ensureDir(moduleWorkflowsDir);
const commandPath = path.join(moduleWorkflowsDir, path.basename(artifact.relativePath));
await this.writeFile(commandPath, artifact.content);
workflowCommandCount++;
}
// Skip workflow-launcher READMEs as they would be treated as slash commands
}
// Generate task and tool commands from manifests (if they exist)
const taskToolGen = new TaskToolCommandGenerator();
const taskToolResult = await taskToolGen.generateTaskToolCommands(projectDir, bmadDir);
console.log(chalk.green(`${this.name} configured:`));
console.log(chalk.dim(` - ${agentCount} agents installed`));
if (workflowCommandCount > 0) {
console.log(chalk.dim(` - ${workflowCommandCount} workflow commands generated`));
}
if (taskToolResult.generated > 0) {
console.log(
chalk.dim(
` - ${taskToolResult.generated} task/tool commands generated (${taskToolResult.tasks} tasks, ${taskToolResult.tools} tools)`,
),
);
}
console.log(chalk.dim(` - Commands directory: ${path.relative(projectDir, bmadCommandsDir)}`));
return {
success: true,
agents: agentCount,
};
}
// Method removed - CLAUDE.md file management left to user
/**
* Read and process file content
*/
async readAndProcess(filePath, metadata) {
const fs = require('fs-extra');
const content = await fs.readFile(filePath, 'utf8');
return this.processContent(content, metadata);
}
/**
* Override processContent to keep {project-root} placeholder
*/
processContent(content, metadata = {}) {
// Use the base class method WITHOUT projectDir to preserve {project-root} placeholder
return super.processContent(content, metadata);
}
/**
* Get agents from source modules (not installed location)
*/
async getAgentsFromSource(sourceDir, selectedModules) {
const fs = require('fs-extra');
const agents = [];
// Add core agents
const corePath = getModulePath('core');
if (await fs.pathExists(path.join(corePath, 'agents'))) {
const coreAgents = await getAgentsFromDir(path.join(corePath, 'agents'), 'core');
agents.push(...coreAgents);
}
// Add module agents
for (const moduleName of selectedModules) {
const modulePath = path.join(sourceDir, moduleName);
const agentsPath = path.join(modulePath, 'agents');
if (await fs.pathExists(agentsPath)) {
const moduleAgents = await getAgentsFromDir(agentsPath, moduleName);
agents.push(...moduleAgents);
}
}
return agents;
}
/**
* Process module injections with pre-collected configuration
*/
async processModuleInjectionsWithConfig(projectDir, bmadDir, options, preCollectedConfig) {
// Get list of installed modules
const modules = options.selectedModules || [];
const { subagentChoices, installLocation } = preCollectedConfig;
// Get the actual source directory (not the installation directory)
await this.processModuleInjectionsInternal({
projectDir,
modules,
handler: 'claude-code',
subagentChoices,
installLocation,
interactive: false,
});
}
/**
* Process Claude Code specific injections for installed modules
* Looks for injections.yaml in each module's claude-code sub-module
*/
async processModuleInjections(projectDir, bmadDir, options) {
// Get list of installed modules
const modules = options.selectedModules || [];
let subagentChoices = null;
let installLocation = null;
// Get the actual source directory (not the installation directory)
const { subagentChoices: updatedChoices, installLocation: updatedLocation } = await this.processModuleInjectionsInternal({
projectDir,
modules,
handler: 'claude-code',
subagentChoices,
installLocation,
interactive: true,
});
if (updatedChoices) {
subagentChoices = updatedChoices;
}
if (updatedLocation) {
installLocation = updatedLocation;
}
}
async processModuleInjectionsInternal({ projectDir, modules, handler, subagentChoices, installLocation, interactive = false }) {
let choices = subagentChoices;
let location = installLocation;
for (const moduleName of modules) {
const configData = await loadModuleInjectionConfig(handler, moduleName);
if (!configData) {
continue;
}
const { config, handlerBaseDir } = configData;
if (interactive) {
console.log(chalk.cyan(`\nConfiguring ${moduleName} ${handler.replace('-', ' ')} features...`));
}
if (interactive && config.subagents && !choices) {
choices = await this.promptSubagentInstallation(config.subagents);
if (choices.install !== 'none') {
const inquirer = require('inquirer');
const locationAnswer = await inquirer.prompt([
{
type: 'list',
name: 'location',
message: 'Where would you like to install Claude Code subagents?',
choices: [
{ name: 'Project level (.claude/agents/)', value: 'project' },
{ name: 'User level (~/.claude/agents/)', value: 'user' },
],
default: 'project',
},
]);
location = locationAnswer.location;
}
}
if (config.injections && choices && choices.install !== 'none') {
for (const injection of config.injections) {
if (shouldApplyInjection(injection, choices)) {
await this.injectContent(projectDir, injection, choices);
}
}
}
if (config.subagents && choices && choices.install !== 'none') {
await this.copySelectedSubagents(projectDir, handlerBaseDir, config.subagents, choices, location || 'project');
}
}
return { subagentChoices: choices, installLocation: location };
}
/**
* Prompt user for subagent installation preferences
*/
async promptSubagentInstallation(subagentConfig) {
const inquirer = require('inquirer');
// First ask if they want to install subagents
const { install } = await inquirer.prompt([
{
type: 'list',
name: 'install',
message: 'Would you like to install Claude Code subagents for enhanced functionality?',
choices: [
{ name: 'Yes, install all subagents', value: 'all' },
{ name: 'Yes, let me choose specific subagents', value: 'selective' },
{ name: 'No, skip subagent installation', value: 'none' },
],
default: 'all',
},
]);
if (install === 'selective') {
// Show list of available subagents with descriptions
const subagentInfo = {
'market-researcher.md': 'Market research and competitive analysis',
'requirements-analyst.md': 'Requirements extraction and validation',
'technical-evaluator.md': 'Technology stack evaluation',
'epic-optimizer.md': 'Epic and story breakdown optimization',
'document-reviewer.md': 'Document quality review',
};
const { selected } = await inquirer.prompt([
{
type: 'checkbox',
name: 'selected',
message: 'Select subagents to install:',
choices: subagentConfig.files.map((file) => ({
name: `${file.replace('.md', '')} - ${subagentInfo[file] || 'Specialized assistant'}`,
value: file,
checked: true,
})),
},
]);
return { install: 'selective', selected };
}
return { install };
}
/**
* Inject content at specified point in file
*/
async injectContent(projectDir, injection, subagentChoices = null) {
const fs = require('fs-extra');
const targetPath = path.join(projectDir, injection.file);
if (await this.exists(targetPath)) {
let content = await fs.readFile(targetPath, 'utf8');
const marker = `<!-- IDE-INJECT-POINT: ${injection.point} -->`;
if (content.includes(marker)) {
let injectionContent = injection.content;
// Filter content if selective subagents chosen
if (subagentChoices && subagentChoices.install === 'selective' && injection.point === 'pm-agent-instructions') {
injectionContent = filterAgentInstructions(injection.content, subagentChoices.selected);
}
content = content.replace(marker, injectionContent);
await fs.writeFile(targetPath, content);
console.log(chalk.dim(` Injected: ${injection.point}${injection.file}`));
}
}
}
/**
* Copy selected subagents to appropriate Claude agents directory
*/
async copySelectedSubagents(projectDir, handlerBaseDir, subagentConfig, choices, location) {
const fs = require('fs-extra');
const os = require('node:os');
// Determine target directory based on user choice
let targetDir;
if (location === 'user') {
targetDir = path.join(os.homedir(), '.claude', 'agents');
console.log(chalk.dim(` Installing subagents globally to: ~/.claude/agents/`));
} else {
targetDir = path.join(projectDir, '.claude', 'agents');
console.log(chalk.dim(` Installing subagents to project: .claude/agents/`));
}
// Ensure target directory exists
await this.ensureDir(targetDir);
const resolvedFiles = await resolveSubagentFiles(handlerBaseDir, subagentConfig, choices);
let copiedCount = 0;
for (const resolved of resolvedFiles) {
try {
const sourcePath = resolved.absolutePath;
const subFolder = path.dirname(resolved.relativePath);
let targetPath;
if (subFolder && subFolder !== '.') {
const targetSubDir = path.join(targetDir, subFolder);
await this.ensureDir(targetSubDir);
targetPath = path.join(targetSubDir, path.basename(resolved.file));
} else {
targetPath = path.join(targetDir, path.basename(resolved.file));
}
await fs.copyFile(sourcePath, targetPath);
console.log(chalk.green(` ✓ Installed: ${subFolder === '.' ? '' : `${subFolder}/`}${path.basename(resolved.file, '.md')}`));
copiedCount++;
} catch (error) {
console.log(chalk.yellow(` ⚠ Error copying ${resolved.file}: ${error.message}`));
}
}
if (copiedCount > 0) {
console.log(chalk.dim(` Total subagents installed: ${copiedCount}`));
}
}
/**
* Install a custom agent launcher for Claude Code
* @param {string} projectDir - Project directory
* @param {string} agentName - Agent name (e.g., "fred-commit-poet")
* @param {string} agentPath - Path to compiled agent (relative to project root)
* @param {Object} metadata - Agent metadata
* @returns {Object|null} Info about created command
*/
async installCustomAgentLauncher(projectDir, agentName, agentPath, metadata) {
const customAgentsDir = path.join(projectDir, this.configDir, this.commandsDir, 'bmad', 'custom', 'agents');
if (!(await this.exists(path.join(projectDir, this.configDir)))) {
return null; // IDE not configured for this project
}
await this.ensureDir(customAgentsDir);
const launcherContent = `---
name: '${agentName}'
description: '${agentName} agent'
---
You must fully embody this agent's persona and follow all activation instructions exactly as specified. NEVER break character until given an exit command.
<agent-activation CRITICAL="TRUE">
1. LOAD the FULL agent file from @${agentPath}
2. READ its entire contents - this contains the complete agent persona, menu, and instructions
3. FOLLOW every step in the <activation> section precisely
4. DISPLAY the welcome/greeting as instructed
5. PRESENT the numbered menu
6. WAIT for user input before proceeding
</agent-activation>
`;
const launcherPath = path.join(customAgentsDir, `${agentName}.md`);
await this.writeFile(launcherPath, launcherContent);
return {
path: launcherPath,
command: `/bmad:custom:agents:${agentName}`,
};
}
}
module.exports = { ClaudeCodeSetup };