@bluefly/openstandardagents 0.1.9

package/.buildkit/agent-first-policy.sh

@@ -0,0 +1,65 @@
+#!/bin/bash
+# Agent-First Policy Enforcement
+# Prevents token waste by encouraging agent usage
+
+set -e
+
+BUILDKIT_CONFIG=".buildkit/branching-workflow.json"
+TOKEN_THRESHOLD=500
+
+# Check if task should use agents
+check_agent_first_policy() {
+    local task_description="$1"
+    local estimated_tokens="$2"
+    
+    echo "🔍 Analyzing task: $task_description"
+    echo "📊 Estimated tokens: $estimated_tokens"
+    
+    if [ "$estimated_tokens" -gt "$TOKEN_THRESHOLD" ]; then
+        echo "⚠️  LARGE TASK DETECTED (>$TOKEN_THRESHOLD tokens)"
+        echo "💡 Recommendation: Use agents to prevent token waste"
+        echo ""
+        echo "Suggested commands:"
+        echo "  buildkit agents list --search '$task_description'"
+        echo "  buildkit agents spawn --type worker --task '$task_description'"
+        echo "  buildkit agents coordinate --multi-agent"
+        echo ""
+        read -p "Continue with manual implementation? (y/N): " -n 1 -r
+        echo
+        if [[ ! $REPLY =~ ^[Yy]$ ]]; then
+            echo "✅ Good choice! Use agents for better token efficiency."
+            exit 1
+        fi
+        echo "⚠️  Proceeding with manual implementation (token inefficient)"
+    fi
+    
+    # Check for repetitive patterns
+    if echo "$task_description" | grep -i -E "(multi-file|database|api.*integration|test.*generation|deployment|refactor)" >/dev/null; then
+        echo "🤖 AGENT-RECOMMENDED: This task type benefits from agent automation"
+        echo "   Consider: buildkit agents spawn --type $(echo "$task_description" | head -c 20)"
+    fi
+}
+
+# Token optimization suggestions
+optimize_tokens() {
+    echo "💰 TOKEN OPTIMIZATION ACTIVE"
+    echo ""
+    echo "Available agent commands:"
+    echo "  buildkit agents list              - Show available agents"
+    echo "  buildkit agents spawn             - Create task-specific agent"
+    echo "  buildkit agents coordinate        - Multi-agent workflows"
+    echo "  buildkit agents status            - Check agent availability"
+    echo ""
+    echo "🎯 Remember: Agents prevent token waste on complex tasks!"
+}
+
+# Main execution
+if [ "$1" = "--check" ]; then
+    check_agent_first_policy "$2" "$3"
+elif [ "$1" = "--optimize" ]; then
+    optimize_tokens
+else
+    echo "Agent-First Policy Enforcement"
+    echo "Usage: $0 --check 'task description' estimated_tokens"
+    echo "       $0 --optimize"
+fi

package/.buildkit/branching-workflow.json

@@ -0,0 +1,41 @@
+{
+  "version": "1.0.0",
+  "enabled": true,
+  "workflow": {
+    "feature_branches_only": true,
+    "ci_driven_development": true,
+    "no_direct_development_commits": true,
+    "main_branch_protection": true
+  },
+  "branch_naming": {
+    "patterns": {
+      "feature": "feature/<scope-kebab-case>",
+      "fix": "fix/<scope-kebab-case>",
+      "chore": "chore/<scope-kebab-case>",
+      "bot": "bot/<agent-name>/<task>"
+    },
+    "validation": {
+      "enforce_naming": true,
+      "reject_version_numbers": true,
+      "require_kebab_case": true
+    }
+  },
+  "ci_rules": {
+    "target_branch": "development",
+    "require_approval": true,
+    "auto_fast_forward": true,
+    "pipeline_success_required": true
+  },
+  "protection": {
+    "block_direct_development_push": true,
+    "block_direct_main_push": true,
+    "quarantine_deleted_branches": true,
+    "backup_on_conflicts": true
+  },
+  "token_optimization": {
+    "agent_first_policy": true,
+    "prevent_manual_coding_over_500_tokens": true,
+    "require_agent_justification": true,
+    "track_token_usage": true
+  }
+}

package/.buildkit/hooks.yaml

@@ -0,0 +1,23 @@
+# PreToolUse Hooks Configuration - Agent BuildKit
+enabled: true
+environment: development
+compliance: [SOC2]
+registered_hooks:
+  - name: command-validation
+    class: CommandValidationHook
+    module: security/CommandValidationHook
+    priority: 100
+    enabled: true
+hooks:
+  security:
+    enabled: true
+    priority: 100
+    config:
+      maxCommandLength: 1000
+      allowedDirectories: [/tmp, /Users/flux423/Sites/LLM/OSSA]
+      blockedDirectories: [/etc, /sys, /proc]
+      blockedCommands: [rm, dd, format]
+      blockedPatterns: ["rm -rf /", "sudo rm", "chmod 777"]
+audit:
+  enabled: true
+  logLevel: info

package/.buildkit/install-buildkit-setup.cjs

@@ -0,0 +1,483 @@
+#!/usr/bin/env node
+
+/**
+ * Enhanced BuildKit Setup Installation for OSSA
+ * Installs comprehensive branching workflow and token optimization
+ */
+
+const fs = require('fs');
+const path = require('path');
+const { execSync } = require('child_process');
+
+// Helper function to ensure directory exists
+function ensureDirSync(dirPath) {
+  if (!fs.existsSync(dirPath)) {
+    fs.mkdirSync(dirPath, { recursive: true });
+  }
+}
+
+// Helper function to write JSON
+function writeJsonSync(filePath, data, options = {}) {
+  const spaces = options.spaces || 0;
+  fs.writeFileSync(filePath, JSON.stringify(data, null, spaces));
+}
+
+console.log('🚀 Installing Agent BuildKit Enhanced Setup on OSSA...\n');
+
+const projectPath = process.cwd();
+console.log(`📁 Project: ${path.basename(projectPath)}`);
+
+try {
+  // 1. Create .buildkit directory
+  const buildkitDir = path.join(projectPath, '.buildkit');
+  ensureDirSync(buildkitDir);
+  console.log('✅ Created .buildkit directory');
+
+  // 2. Install branching workflow specification
+  console.log('🌿 Installing branching & CI workflow enforcement...');
+  
+  const branchingSpec = {
+    version: '1.0.0',
+    enabled: true,
+    workflow: {
+      feature_branches_only: true,
+      ci_driven_development: true,
+      no_direct_development_commits: true,
+      main_branch_protection: true
+    },
+    branch_naming: {
+      patterns: {
+        feature: 'feature/<scope-kebab-case>',
+        fix: 'fix/<scope-kebab-case>',
+        chore: 'chore/<scope-kebab-case>',
+        bot: 'bot/<agent-name>/<task>'
+      },
+      validation: {
+        enforce_naming: true,
+        reject_version_numbers: true,
+        require_kebab_case: true
+      }
+    },
+    ci_rules: {
+      target_branch: 'development',
+      require_approval: true,
+      auto_fast_forward: true,
+      pipeline_success_required: true
+    },
+    protection: {
+      block_direct_development_push: true,
+      block_direct_main_push: true,
+      quarantine_deleted_branches: true,
+      backup_on_conflicts: true
+    },
+    token_optimization: {
+      agent_first_policy: true,
+      prevent_manual_coding_over_500_tokens: true,
+      require_agent_justification: true,
+      track_token_usage: true
+    }
+  };
+  
+  writeJsonSync(
+    path.join(buildkitDir, 'branching-workflow.json'),
+    branchingSpec,
+    { spaces: 2 }
+  );
+  console.log('✅ Installed branching workflow specification');
+
+  // 3. Install GitLab CI template with golden workflow
+  console.log('🏗️ Installing GitLab CI with golden workflow integration...');
+  
+  const gitlabCITemplate = `# GitLab CI - Agent BuildKit Enhanced Pipeline
+# Uses golden workflow component for enterprise-grade CI/CD
+
+include:
+  - component: gitlab.bluefly.io/llm/gitlab_components/workflow/golden@v0.1.0
+    inputs:
+      project_name: "\${CI_PROJECT_NAME}"
+      enable_auto_flow: true
+      enable_comprehensive_testing: true
+      enable_security_scanning: true
+      test_coverage_threshold: 80
+      enable_agent_buildkit_validation: true
+      branch_protection_enabled: true
+
+variables:
+  # Agent BuildKit Configuration
+  BUILDKIT_ENABLED: "true"
+  BUILDKIT_TOKEN_OPTIMIZATION: "true"
+  BUILDKIT_BRANCH_ENFORCEMENT: "true"
+  
+  # Branch Protection
+  DEVELOPMENT_BRANCH_PROTECTED: "true"
+  MAIN_BRANCH_PROTECTED: "true"
+  FEATURE_BRANCH_REQUIRED: "true"
+
+stages:
+  - validate
+  - test
+  - security
+  - build
+  - deploy
+
+# Agent BuildKit Branch Validation
+buildkit:branch-validation:
+  stage: validate
+  image: node:18-alpine
+  script:
+    - npm install -g @bluefly/agent-build-kit
+    - buildkit hooks validate git "branch --show-current"
+    - |
+      BRANCH=$(git branch --show-current)
+      case "$BRANCH" in
+        feature/*|fix/*|chore/*|bot/*)
+          echo "✅ Valid branch naming: $BRANCH"
+          ;;
+        development|main)
+          echo "❌ Direct work on $BRANCH is not allowed"
+          exit 1
+          ;;
+        *)
+          echo "❌ Invalid branch naming: $BRANCH"
+          echo "ℹ️  Use: feature/<scope>, fix/<scope>, chore/<scope>, or bot/<agent>/<task>"
+          exit 1
+          ;;
+      esac
+  rules:
+    - if: $CI_PIPELINE_SOURCE == "merge_request_event"
+    - if: $CI_COMMIT_BRANCH != "main" && $CI_COMMIT_BRANCH != "development"
+
+# Token Optimization Check
+buildkit:token-optimization:
+  stage: validate
+  image: node:18-alpine
+  script:
+    - npm install -g @bluefly/agent-build-kit
+    - buildkit agents list --check-availability
+    - echo "🤖 Verifying agent-first policy compliance"
+    - |
+      # Check if large changes should use agents instead
+      CHANGED_LINES=$(git diff --stat HEAD~1 | tail -1 | grep -o '[0-9]\\+ insertions' | grep -o '[0-9]\\+' || echo "0")
+      if [ "$CHANGED_LINES" -gt 500 ]; then
+        echo "⚠️  Large changeset detected ($CHANGED_LINES lines)"
+        echo "💡 Consider using buildkit agents for token optimization"
+        echo "   Run: buildkit agents spawn --type worker --task 'large-refactor'"
+      fi
+  allow_failure: true
+  rules:
+    - if: $CI_PIPELINE_SOURCE == "merge_request_event"
+
+# Prevent AI Attribution in Commits
+buildkit:employment-protection:
+  stage: validate
+  image: alpine/git
+  script:
+    - |
+      echo "🔍 Checking commits for AI attribution..."
+      if git log --oneline -10 | grep -i -E "(claude|🤖|generated.*with|anthropic)"; then
+        echo "❌ AI attribution detected in commit messages"
+        echo "🚨 EMPLOYMENT RISK: Remove AI references from commit history"
+        exit 1
+      fi
+      echo "✅ No AI attribution found"
+  rules:
+    - if: $CI_PIPELINE_SOURCE == "merge_request_event"
+`;
+  
+  fs.writeFileSync(path.join(projectPath, '.gitlab-ci.yml'), gitlabCITemplate);
+  console.log('✅ Installed GitLab CI with golden workflow integration');
+
+  // 4. Create agent-first policy enforcement script
+  console.log('🤖 Installing agent-first policy enforcement...');
+  
+  const agentFirstScript = `#!/bin/bash
+# Agent-First Policy Enforcement
+# Prevents token waste by encouraging agent usage
+
+set -e
+
+BUILDKIT_CONFIG=".buildkit/branching-workflow.json"
+TOKEN_THRESHOLD=500
+
+# Check if task should use agents
+check_agent_first_policy() {
+    local task_description="$1"
+    local estimated_tokens="$2"
+    
+    echo "🔍 Analyzing task: $task_description"
+    echo "📊 Estimated tokens: $estimated_tokens"
+    
+    if [ "$estimated_tokens" -gt "$TOKEN_THRESHOLD" ]; then
+        echo "⚠️  LARGE TASK DETECTED (>$TOKEN_THRESHOLD tokens)"
+        echo "💡 Recommendation: Use agents to prevent token waste"
+        echo ""
+        echo "Suggested commands:"
+        echo "  buildkit agents list --search '$task_description'"
+        echo "  buildkit agents spawn --type worker --task '$task_description'"
+        echo "  buildkit agents coordinate --multi-agent"
+        echo ""
+        read -p "Continue with manual implementation? (y/N): " -n 1 -r
+        echo
+        if [[ ! $REPLY =~ ^[Yy]$ ]]; then
+            echo "✅ Good choice! Use agents for better token efficiency."
+            exit 1
+        fi
+        echo "⚠️  Proceeding with manual implementation (token inefficient)"
+    fi
+    
+    # Check for repetitive patterns
+    if echo "$task_description" | grep -i -E "(multi-file|database|api.*integration|test.*generation|deployment|refactor)" >/dev/null; then
+        echo "🤖 AGENT-RECOMMENDED: This task type benefits from agent automation"
+        echo "   Consider: buildkit agents spawn --type $(echo "$task_description" | head -c 20)"
+    fi
+}
+
+# Token optimization suggestions
+optimize_tokens() {
+    echo "💰 TOKEN OPTIMIZATION ACTIVE"
+    echo ""
+    echo "Available agent commands:"
+    echo "  buildkit agents list              - Show available agents"
+    echo "  buildkit agents spawn             - Create task-specific agent"
+    echo "  buildkit agents coordinate        - Multi-agent workflows"
+    echo "  buildkit agents status            - Check agent availability"
+    echo ""
+    echo "🎯 Remember: Agents prevent token waste on complex tasks!"
+}
+
+# Main execution
+if [ "$1" = "--check" ]; then
+    check_agent_first_policy "$2" "$3"
+elif [ "$1" = "--optimize" ]; then
+    optimize_tokens
+else
+    echo "Agent-First Policy Enforcement"
+    echo "Usage: $0 --check 'task description' estimated_tokens"
+    echo "       $0 --optimize"
+fi
+`;
+  
+  fs.writeFileSync(path.join(buildkitDir, 'agent-first-policy.sh'), agentFirstScript);
+  fs.chmodSync(path.join(buildkitDir, 'agent-first-policy.sh'), '755');
+  console.log('✅ Installed agent-first policy enforcement script');
+
+  // 5. Install enhanced lefthook.yml with employment protection
+  console.log('🛡️ Installing employment protection (lefthook.yml)...');
+  
+  const lefthookConfig = `# GitHook Management - Agent BuildKit Employment Protection
+pre-push:
+  commands:
+    block-claude-on-development:
+      run: |
+        BRANCH=$(git branch --show-current)
+        if [ "$BRANCH" = "development" ]; then
+          echo "❌ BLOCKED: Claude Code is not allowed to push to development branch"
+          echo "🚨 EMPLOYMENT PROTECTION: Development branch is for human developers only"
+          echo "ℹ️  Use: git checkout -b feature/your-feature"
+          exit 1
+        fi
+
+commit-msg:
+  commands:
+    block-claude-attribution:
+      run: |
+        MSG_FILE={1}
+        if grep -i -E "(claude|🤖|generated with.*claude|co-authored-by.*claude|anthropic)" "$MSG_FILE"; then
+          echo "❌ COMMIT REJECTED: Claude/AI attribution detected"
+          echo "🚨 EMPLOYMENT RISK: Commits must appear to be from human developers"
+          echo "ℹ️  Remove AI attribution and commit as yourself"
+          exit 1
+        fi
+
+pre-commit:
+  commands:
+    prevent-ai-signatures:
+      run: |
+        if git diff --cached | grep -i -E "(generated.*claude|ai.*generated|anthropic|claude.*code)"; then
+          echo "❌ COMMIT BLOCKED: AI signature detected in code"
+          echo "🚨 EMPLOYMENT PROTECTION: Remove AI references from code"
+          exit 1
+        fi
+`;
+  
+  fs.writeFileSync(path.join(projectPath, 'lefthook.yml'), lefthookConfig);
+  console.log('✅ Installed branch protection (lefthook.yml)');
+
+  // 6. Create comprehensive CLAUDE.md with workflow specification
+  console.log('📋 Creating comprehensive CLAUDE.md with workflow specification...');
+  
+  const claudeConfig = `# CLAUDE.md - OSSA Project Configuration
+
+## 🚨 CRITICAL: Branching & CI Workflow Specification
+
+### 1. Feature Branches Only (MANDATORY)
+- All new work must start from the latest \`development\` branch
+- Branch names use the format:
+  - \`feature/<scope-kebab-case>\` for new capabilities
+  - \`fix/<scope-kebab-case>\` for bug fixes  
+  - \`chore/<scope-kebab-case>\` for non-feature work (docs, refactors, CI config)
+  - \`bot/<agent-name>/<task>\` for AI bot work
+
+### 2. CI-Driven Development Branch (ENFORCED)
+- Merge requests target \`development\`, never \`main\` directly
+- CI/CD enforces: linting, type checks, tests, coverage thresholds
+- Approval requirements defined in CODEOWNERS
+- Automatic fast-forward merge after approval and pipeline success
+- **No manual merges to development are allowed. Only CI merges.**
+
+### 3. AI Bots Sync from Development
+- Bots always pull from \`development\` before creating new branches
+- Bot branches follow naming convention: \`bot/<agent-name>/<task>\`
+- CI ensures bot-created branches are isolated and only merged via standard MR flow
+
+### 4. No Direct Development Commits (BLOCKED)
+- Direct pushes to \`development\` are blocked at Git server level
+- Developers must create feature/fix branches and submit merge requests
+- Development is managed exclusively through automation and CI
+
+### 5. Main Branch Protection (AUTOMATED)
+- \`main\` only accepts merges from \`development\` via release workflow
+- Releases follow semantic versioning:
+  - \`feat:\` → minor version bump
+  - \`fix:\` → patch version bump  
+  - \`BREAKING CHANGE:\` → major version bump
+- \`main\` is always deployable and tagged automatically with semantic-release
+
+### 6. Safety & Recovery Rules (AUTOMATIC)
+- Backup branches auto-created by CI on merge conflicts or failed automations
+- Deleted branches quarantined to \`__DELETE_LATER/\` until cleanup job runs
+
+## 💰 TOKEN OPTIMIZATION (MANDATORY)
+
+### Agent-First Policy - STOP WASTING TOKENS
+Before writing ANY code, you MUST check for agents:
+
+\`\`\`bash
+# Step 1: MANDATORY token estimation
+echo "Task: [description]"
+echo "Estimated tokens: [amount]" 
+echo "Is this > 500 tokens? [yes/no]"
+echo "Is this repetitive? [yes/no]"
+
+# Step 2: MANDATORY agent check
+buildkit agents list
+buildkit agents search --tag [keyword]
+buildkit agents search --description "[task-description]"
+
+# Step 3: Decision matrix
+if [estimated_tokens > 500] || [is_repetitive]; then
+    echo "MUST use agents - token waste prevention"
+    buildkit agents spawn --type worker --task "[description]"
+else
+    echo "Manual coding approved - justification: [reason]"
+fi
+\`\`\`
+
+### MANDATORY Agent Spawning for These Tasks:
+- **Multi-file operations** (3+ files) → ALWAYS spawn agents
+- **Database operations** → ALWAYS use database agents
+- **API integrations** → ALWAYS use integration agents
+- **Testing workflows** → ALWAYS use test agents
+- **Deployment tasks** → ALWAYS use deployment agents
+- **Documentation generation** → ALWAYS use doc agents
+- **Code analysis** (linting, security) → ALWAYS use analysis agents
+- **Refactoring tasks** → ALWAYS use refactoring agents
+
+### Token Waste Prevention Rules:
+- ❌ **NEVER manually code tasks > 500 tokens without checking agents**
+- ❌ **NEVER repeat similar tasks without creating/using agents**
+- ❌ **NEVER ignore existing agent capabilities**
+- ❌ **NEVER write multi-step workflows manually**
+- ❌ **NEVER generate boilerplate code manually**
+
+## 🛡️ EMPLOYMENT PROTECTION ACTIVE
+
+### Critical Restrictions:
+- **Claude Code is BLOCKED from development branch**
+- **All AI assistants must work on feature branches only**
+- **Commits must NOT contain AI attribution**
+- **No Claude/Anthropic references in code or commits**
+
+### Required Workflow:
+\`\`\`bash
+# ALWAYS start from development
+git checkout development
+git pull origin development
+
+# Create feature branch (REQUIRED)
+git checkout -b feature/your-description
+
+# Work on feature branch only
+# ... make changes ...
+
+# Commit as human developer (NO AI attribution)
+git commit -m "feat: implement new feature"
+
+# Push feature branch
+git push origin feature/your-description
+\`\`\`
+
+### Blocked Actions:
+- ❌ Direct work on development branch
+- ❌ AI attribution in commits (\`🤖 Generated with Claude\`)
+- ❌ Claude Code signatures in code
+- ❌ Anthropic references in commits
+- ❌ Co-authored-by Claude entries
+
+## 🎯 Success Criteria
+
+An AI assistant is properly configured when:
+- ✅ Always checks agent-buildkit before manual coding
+- ✅ Uses correct branch naming patterns
+- ✅ Works only from development branch locally
+- ✅ Implements golden workflow CI/CD
+- ✅ Follows TDD and API-first principles  
+- ✅ Never creates ad-hoc scripts outside agent-buildkit
+
+Generated by Agent BuildKit Enhanced Setup
+Protected by comprehensive governance system
+`;
+  
+  fs.writeFileSync(path.join(projectPath, 'CLAUDE.md'), claudeConfig);
+  console.log('✅ Created comprehensive CLAUDE.md with workflow specification');
+
+  // 7. Update .gitignore
+  const gitignorePath = path.join(projectPath, '.gitignore');
+  if (fs.existsSync(gitignorePath)) {
+    const gitignoreContent = fs.readFileSync(gitignorePath, 'utf8');
+    if (!gitignoreContent.includes('.buildkit/cache')) {
+      fs.appendFileSync(gitignorePath, '\n# Agent BuildKit\n.buildkit/cache/\n.buildkit/logs/\n');
+      console.log('✅ Updated .gitignore for Agent BuildKit');
+    }
+  }
+
+  // 8. Try to install lefthook if available
+  try {
+    console.log('🔗 Attempting to install lefthook hooks...');
+    execSync('npx lefthook install', { stdio: 'ignore' });
+    console.log('✅ Lefthook hooks activated');
+  } catch {
+    console.log('⚠️  Lefthook not available - install with: npm install -g lefthook');
+  }
+
+  // Final summary
+  console.log('\n📊 Installation Summary:');
+  console.log('  ✅ Branching workflow specification installed');
+  console.log('  ✅ GitLab CI golden workflow template installed');
+  console.log('  ✅ Agent-first policy enforcement installed');
+  console.log('  ✅ Employment protection activated');
+  console.log('  ✅ Comprehensive CLAUDE.md created');
+  console.log('  ✅ Token optimization enabled');
+
+  console.log('\n🤖 Token Optimization Enabled:');
+  console.log('  • Use "buildkit agents list" to check available agents');
+  console.log('  • Use "buildkit agents spawn" for complex tasks');
+  console.log('  • Agent-first policy prevents token waste');
+  console.log('  • Branching workflow enforces CI-driven development');
+
+  console.log('\n🎉 Agent BuildKit Enhanced Setup complete! Ready for enterprise development.');
+
+} catch (error) {
+  console.error('❌ Setup failed:', error.message);
+  process.exit(1);
+}