@ema.co/mcp-toolkit 2026.1.24 → 2026.1.26-3

package/resources/templates/dashboard-ai/persona-config.json

@@ -1,180 +0,0 @@
-{
-  "$schema": "https://ema.co/schemas/persona-config-dashboard.json",
-  "$comment": "Dashboard AI Persona Configuration - Fill in all {placeholder} values",
-  
-  "personaType": "dashboard",
-  "projectType": 2,
-  "name": "{Persona Name}",
-  "description": "{Brief description of what this Dashboard AI does}",
-  
-  "dashboardSettings": {
-    "name": "{Display name}",
-    "description": "{What this dashboard processes}",
-    
-    "inputSchema": {
-      "fields": [
-        {
-          "name": "{field1}",
-          "type": "file",
-          "required": true,
-          "description": "{Primary document to process}",
-          "acceptedTypes": ["application/pdf", "image/*"]
-        },
-        {
-          "name": "{field2}",
-          "type": "string",
-          "required": false,
-          "description": "{Optional text field}"
-        },
-        {
-          "name": "{field3}",
-          "type": "select",
-          "required": true,
-          "options": ["{option1}", "{option2}", "{option3}"],
-          "description": "{Category selection}"
-        }
-      ]
-    },
-    
-    "batchSettings": {
-      "batchSize": 50,
-      "$comment_batchSize": "Number of items per batch (1-100)",
-      "timeout": 60,
-      "$comment_timeout": "Seconds per item before timeout",
-      "parallelExecution": true,
-      "$comment_parallelExecution": "Process items in parallel"
-    },
-    
-    "outputSchema": {
-      "fields": [
-        {
-          "name": "status",
-          "type": "enum",
-          "values": ["approved", "pending_review", "rejected"]
-        },
-        {
-          "name": "extracted_data",
-          "type": "object",
-          "description": "Extracted entity values"
-        },
-        {
-          "name": "validation_errors",
-          "type": "array",
-          "description": "List of validation failures"
-        },
-        {
-          "name": "confidence_score",
-          "type": "number",
-          "description": "0-1 confidence in extraction"
-        }
-      ]
-    }
-  },
-  
-  "entityExtraction": {
-    "columns": [
-      {
-        "name": "{Entity 1}",
-        "type": "string",
-        "description": "{What to extract}",
-        "examples": ["{example1}", "{example2}"]
-      },
-      {
-        "name": "{Entity 2}",
-        "type": "number",
-        "description": "{Numeric value to extract}",
-        "examples": ["{example1}", "{example2}"]
-      },
-      {
-        "name": "{Entity 3}",
-        "type": "date",
-        "description": "{Date to extract}",
-        "format": "YYYY-MM-DD"
-      }
-    ],
-    "$comment": "Define columns for entity_extraction_with_documents action"
-  },
-  
-  "ruleValidation": {
-    "rules": [
-      {
-        "name": "{Rule 1}",
-        "description": "{What this rule checks}",
-        "condition": "{Entity 1} is not empty",
-        "severity": "error"
-      },
-      {
-        "name": "{Rule 2}",
-        "description": "{What this rule checks}",
-        "condition": "{Entity 2} > 0",
-        "severity": "warning"
-      }
-    ],
-    "$comment": "Define rules for rule_validation_with_documents action"
-  },
-  
-  "approvalRouting": {
-    "rules": [
-      {
-        "name": "Auto-approve",
-        "condition": "confidence_score >= 0.95 AND validation_errors.length == 0",
-        "action": "auto_approve"
-      },
-      {
-        "name": "Manager Review",
-        "condition": "{amount} > 10000",
-        "action": "route_to_manager",
-        "assignee": "{manager_email}"
-      },
-      {
-        "name": "Reject",
-        "condition": "validation_errors contains 'critical'",
-        "action": "reject"
-      }
-    ],
-    "sla": {
-      "standard": "24h",
-      "urgent": "4h",
-      "reminderAt": "50%",
-      "escalateAt": "100%"
-    }
-  },
-  
-  "fileUpload": {
-    "localFiles": [],
-    "tags": ["{default-tag}"],
-    "useChunking": true,
-    "acceptedMimeTypes": [
-      "application/pdf",
-      "image/png",
-      "image/jpeg",
-      "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"
-    ]
-  },
-  
-  "dataProtection": {
-    "protectedClasses": [
-      "ssn",
-      "credit_card",
-      "account_number"
-    ]
-  },
-  
-  "fusionModel": {
-    "allModels": true,
-    "selectedModels": []
-  },
-  
-  "notifications": {
-    "onSuccess": {
-      "enabled": true,
-      "recipients": ["{email}"],
-      "template": "Processing complete: {count} items processed"
-    },
-    "onFailure": {
-      "enabled": true,
-      "recipients": ["{email}"],
-      "template": "Processing failed: {error}"
-    }
-  }
-}

package/resources/templates/demo-scenarios/README.md

@@ -1,63 +0,0 @@
-# Demo Scenarios
-
-Pre-built demo data scenarios for reliable AI Employee demonstrations.
-
-## Available Scenarios
-
-| Scenario | Use Case | Persona Types |
-|----------|----------|---------------|
-| `sales-sdr` | Sales Development Representative | Voice, Chat |
-| `support-tier1` | Customer Support Tier 1 | Voice, Chat |
-| `hr-assistant` | HR Policy & Benefits Assistant | Chat, Voice |
-
-## Usage
-
-Generate a demo kit:
-
-```
-demo(mode="kit", persona_id="...", scenario="sales-sdr")
-```
-
-Validate demo readiness:
-
-```
-demo(mode="validate_kit", persona_id="...")
-```
-
-## What's Generated
-
-Each demo kit includes:
-
-1. **KB Documents** - Markdown files for the knowledge base with demo-ready data
-2. **Demo Script** - Questions to ask with expected answers
-3. **Fixed Response Fallbacks** - Workflow nodes for guaranteed responses
-4. **Validation Queries** - Test queries to verify before the demo
-
-## Customization
-
-Add custom Q&A pairs:
-
-```
-demo(mode="kit", persona_id="...", scenario="sales-sdr", custom_qa=[
-  {"question": "Tell me about feature X", "answer": "Feature X does..."}
-])
-```
-
-## Creating Custom Scenarios
-
-Create a JSON file in this directory with the following structure:
-
-```json
-{
-  "id": "my-scenario",
-  "name": "My Custom Scenario",
-  "description": "Description",
-  "persona_types": ["voice", "chat"],
-  "tags": ["tag1", "tag2"],
-  "intents": [...],
-  "entities": [...],
-  "qa_pairs": [...]
-}
-```
-
-See existing scenarios for complete examples.

package/resources/templates/demo-scenarios/test-published-package.md

@@ -1,116 +0,0 @@
-# Test Published Package: @ema.co/mcp-toolkit
-
-Use this prompt to verify the published npm package works correctly.
-
-## Quick Test (Copy & Run)
-
-```bash
-# Create temp directory
-cd /tmp && rm -rf mcp-test && mkdir mcp-test && cd mcp-test
-
-# Initialize and install
-npm init -y && npm install @ema.co/mcp-toolkit@latest
-
-# Create test file
-cat > test.mjs << 'EOF'
-import * as sdk from '@ema.co/mcp-toolkit/sdk';
-
-console.log('=== @ema.co/mcp-toolkit Test Suite ===\n');
-
-// 1. Version check
-const pkg = await import('@ema.co/mcp-toolkit/sdk');
-console.log(`SDK exports: ${Object.keys(pkg).length} items\n`);
-
-// 2. Core classes
-console.log('Core Classes:');
-console.log(`  EmaClient: ${typeof sdk.EmaClient === 'function' ? '✅' : '❌'}`);
-console.log(`  APISchemaRegistry: ${typeof sdk.APISchemaRegistry === 'function' ? '✅' : '❌'}`);
-console.log(`  StateStore: ${typeof sdk.StateStore === 'function' ? '✅' : '❌'}`);
-
-// 3. Workflow generation
-console.log('\nWorkflow Generation:');
-console.log(`  parseNaturalLanguage: ${typeof sdk.parseNaturalLanguage === 'function' ? '✅' : '❌'}`);
-console.log(`  needsLLMGeneration: ${typeof sdk.needsLLMGeneration === 'function' ? '✅' : '❌'}`);
-console.log(`  generateWorkflow: ${typeof sdk.generateWorkflow === 'function' ? '✅' : '❌'}`);
-
-// 4. Validation
-console.log('\nValidation:');
-console.log(`  validateWorkflowSpec: ${typeof sdk.validateWorkflowSpec === 'function' ? '✅' : '❌'}`);
-
-// 5. Functional test
-console.log('\nFunctional Test:');
-const intent = sdk.parseNaturalLanguage('Customer support bot that searches knowledge base and sends emails');
-console.log(`  Parse intent: ${intent.intents?.length > 0 ? '✅' : '❌'}`);
-console.log(`  Intents found: ${JSON.stringify(intent.intents?.map(i => i.name))}`);
-
-const needsLLM = sdk.needsLLMGeneration(intent);
-console.log(`  Needs LLM generation: ${needsLLM}`);
-
-if (needsLLM) {
-  const result = sdk.generateWorkflow(intent, [], []);
-  console.log(`  LLM prompt generated: ${result.llm_prompt ? '✅' : '❌'}`);
-}
-
-console.log('\n=== Test Complete ===');
-EOF
-
-# Run test
-node test.mjs
-```
-
-## Expected Output
-
-```
-=== @ema.co/mcp-toolkit Test Suite ===
-
-SDK exports: 119+ items
-
-Core Classes:
-  EmaClient: ✅
-  APISchemaRegistry: ✅
-  StateStore: ✅
-
-Workflow Generation:
-  parseNaturalLanguage: ✅
-  needsLLMGeneration: ✅
-  generateWorkflow: ✅
-
-Validation:
-  validateWorkflowSpec: ✅
-
-Functional Test:
-  Parse intent: ✅
-  Intents found: ["Customer Support","Email Automation"]
-  Needs LLM generation: true
-  LLM prompt generated: ✅
-
-=== Test Complete ===
-```
-
-## Import Paths
-
-| Path | Contents |
-|------|----------|
-| `@ema.co/mcp-toolkit/sdk` | SDK classes and functions |
-| `@ema.co/mcp-toolkit/mcp` | MCP server components |
-| `@ema.co/mcp-toolkit` | CLI entry point (not for programmatic use) |
-
-## MCP Server Test
-
-To test the MCP server:
-
-```bash
-# Start MCP server (requires EMA_API_TOKEN)
-export EMA_API_TOKEN="your-token"
-export EMA_TENANT_ID="your-tenant"
-npx -y @ema.co/mcp-toolkit@latest
-
-# Or with npx directly
-EMA_API_TOKEN="..." EMA_TENANT_ID="..." npx -y @ema.co/mcp-toolkit@latest
-```
-
-## Version Check
-
-```bash
-npm view @ema.co/mcp-toolkit version
-```

package/resources/templates/document-gen-ai/README.md

@@ -1,132 +0,0 @@
-# Document Generation AI - Resume Analysis Report Generator
-
-This template provides a document generation persona for creating comprehensive resume analysis reports.
-
-## Overview
-
-The Resume Analysis Report Generator helps HR teams and hiring managers quickly evaluate candidates through AI-powered resume analysis. It extracts structured data from resumes, validates candidates against configurable criteria, and generates detailed assessment reports.
-
-## Report Types
-
-### 1. Candidate Assessment Report
-Comprehensive analysis of a single candidate including:
-- Executive summary with overall recommendation
-- Skills assessment (technical and soft skills)
-- Experience analysis and career trajectory
-- Education and credentials review
-- Fit scoring (technical, cultural, overall on 1-10 scale)
-- Concerns, red flags, and recommendations
-
-### 2. Skills Gap Analysis
-Compare candidate qualifications against job requirements:
-- Required skills inventory from job description
-- Candidate skills inventory from resume
-- Gap identification matrix
-- Severity assessment (critical, moderate, minor)
-- Training and development recommendations
-- Time-to-proficiency estimates
-
-### 3. Candidate Comparison Report
-Side-by-side comparison of multiple candidates:
-- Summary comparison table
-- Individual candidate profiles
-- Skills matrix comparison
-- Experience level comparison
-- Ranked recommendations
-- Interview priority order
-
-## Key Features
-
-### Entity Extraction
-Automatically extracts structured data from resumes:
-- Candidate name and contact information
-- Total years of experience
-- Current/most recent role
-- Education history (degrees, institutions, fields)
-- Technical skills inventory
-- Professional certifications
-- Complete work history with achievements
-
-### Rule Validation
-Validates candidates against configurable hiring criteria:
-- Minimum experience requirements
-- Education standards
-- Communication skills evidence
-- Career progression patterns
-- Employment gap analysis
-
-### Fit Scoring
-Provides quantified assessment across dimensions:
-- Technical Fit (40% weight) - Skills match to requirements
-- Experience Fit (35% weight) - Relevance and depth
-- Cultural Fit (25% weight) - Values and team alignment
-- Overall Recommendation: Strong Yes / Yes / Maybe / No
-
-## Configuration
-
-### Required Inputs
-- Resume documents (PDF, DOCX)
-
-### Optional Inputs
-- Job description for comparison analysis
-- Analysis instructions (focus areas, scoring criteria)
-- Candidate tags for categorization
-
-### Data Protection
-By default, the following are redacted from LLM processing:
-- Phone numbers
-- Email addresses
-- Physical addresses
-- Social security numbers
-
-## Usage with MCP
-
-```typescript
-// Create persona from template
-persona(input="Create a Resume Analysis Report Generator for our Engineering team", type="dashboard")
-
-// Upload resume for analysis
-data(persona_id="...", action="upload", file="candidate_resume.pdf")
-
-// Generate assessment report
-// (Done through Ema dashboard UI)
-```
-
-## Customization
-
-### Adding Extraction Fields
-Modify the `extraction_columns` in the workflow definition to extract additional data points:
-- Languages spoken
-- Salary expectations
-- Notice period
-- Remote work preferences
-- Portfolio/GitHub links
-
-### Custom Validation Rules
-Add rules to the `all_rulesets` configuration:
-- Industry-specific requirements
-- Skill-based criteria
-- Certification requirements
-- Location preferences
-
-### Report Sections
-Configure `sectionConfigWidget` to customize report structure:
-- Add/remove sections
-- Reorder sections
-- Add custom section types
-
-## File Support
-
-Supported resume formats:
-- PDF
-- DOCX
-- DOC
-- TXT
-
-## Best Practices
-
-1. **Upload job descriptions** - Enables more accurate fit scoring and gap analysis
-2. **Customize validation rules** - Tailor criteria to your specific role requirements
-3. **Use tags** - Categorize candidates by role type, experience level, location
-4. **Review extraction accuracy** - Validate extracted data on first few resumes
-5. **Iterate on prompts** - Refine analysis instructions based on report quality