Part of VISHKAR MCP Ecosystem← Back to Ecosystem Hub

StoryCrafter MCP

AI-Powered Backlog Generation - Epics & Stories from VISHKAR Context

4 ToolsFast Epics (15-20s)Detailed Storiesv1.0.0

Role in the Ecosystem

StoryCrafter MCP is the AI Backlog Generator of the VISHKAR ecosystem. It transforms project context into well-structured epics and user stories with acceptance criteria, technical tasks, and estimates.

  • Fast epic generation (15-20 seconds)
  • Detailed story generation (3-5 minutes)
  • Iterative refinement with feedback
  • Ready for JIRA import

Quick Example

# Generate epics from context
curl -X POST \
  https://storycrafter-mcp.vercel.app/api/mcp \
  -H "X-API-Key: pk_your_key" \
  -d '{
    "method": "tools/call",
    "params": {
      "tool": "generate_epics",
      "arguments": {
        "project_context": {...}
      }
    }
  }'

Ecosystem Architecture

┌─────────────────────────────────────────────────────────────────┐
│                     VISHKAR AI Agent                             │
│          (Collects project_summary + final_decisions)            │
└──────────────────────────────┬──────────────────────────────────┘
                               │
                               │ project_context
                               ▼
┌───────────────────────────────────────────────────────────────────┐
│                      STORYCRAFTER MCP                             │ ◄── YOU ARE HERE
│   ┌─────────────────────────────────────────────────────────┐    │
│   │  Phase 1: generate_epics (Claude Sonnet - 15-20s)       │    │
│   │  Phase 2: generate_stories (GPT - 3-5min)               │    │
│   │  Refinement: regenerate_epic, regenerate_story          │    │
│   └─────────────────────────────────────────────────────────┘    │
└──────────────────────────────┬──────────────────────────────────┘
                               │
                               │ Epics & Stories JSON
                               ▼
┌───────────────────────────────────────────────────────────────────┐
│                         JIRA MCP                                   │
│              bulk_create_issues → Import to JIRA                   │
└───────────────────────────────────────────────────────────────────┘

Workflow:
1. VISHKAR collects project requirements
2. StoryCrafter generates epics (fast, high-level)
3. User reviews and refines epics
4. StoryCrafter generates detailed stories
5. User reviews and refines stories
6. Output ready for JIRA import

Two-Phase Generation Workflow

Phase 1: Epic Generation

⚡ Fast

Quick generation of 5-8 high-level epics from project context.

Duration:

15-20 seconds

Model:

Claude Sonnet

Cost:

~$0.15

Output:

5-8 Epics

Tools: generate_epics, regenerate_epic

Phase 2: Story Generation

🔄 Detailed

Comprehensive user stories with acceptance criteria and technical tasks.

Duration:

3-5 minutes

Model:

GPT (128K output)

Cost:

~$0.29

Output:

Full backlog

Tools: generate_stories, regenerate_story

Available Tools

ToolDescriptionSpeed
generate_epicsGenerate 5-8 high-level epics from context15-20s
generate_storiesGenerate detailed stories for epics3-5min
regenerate_epicRefine single epic based on feedback20-30s
regenerate_storyRefine single story based on feedback1-2min

Input & Output

Input: VISHKAR Project Context

{
  "project_context": {
    "project_summary": {
      "name": "TaskMaster",
      "description": "Task app with offline",
      "target_users": "Remote teams"
    },
    "final_decisions": {
      "technical_stack": "React Native",
      "timeline": "4 weeks",
      "priorities": ["Offline", "Sync"]
    }
  }
}

Output: Structured Backlog

{
  "backlog": {
    "epics": [{
      "id": "EPIC-1",
      "title": "User Authentication",
      "stories": [{
        "id": "STORY-1",
        "title": "User Registration",
        "acceptance_criteria": [...],
        "technical_tasks": [...],
        "story_points": 5,
        "estimated_hours": 12
      }]
    }],
    "summary": {
      "total_epics": 8,
      "total_stories": 36
    }
  }
}

Full Workflow Example

Step 1: Generate Epics

POST /api/mcp
{
  "method": "tools/call",
  "params": {
    "tool": "generate_epics",
    "arguments": {
      "project_context": {
        "project_summary": { "name": "TaskMaster", ... },
        "final_decisions": { "timeline": "4 weeks", ... }
      }
    }
  }
}

Step 2: Refine Epic (if needed)

POST /api/mcp
{
  "method": "tools/call",
  "params": {
    "tool": "regenerate_epic",
    "arguments": {
      "project_context": { ... },
      "epic": { "id": "EPIC-1", "title": "...", ... },
      "user_feedback": "Focus more on OAuth integration"
    }
  }
}

Step 3: Generate Stories

POST /api/mcp
{
  "method": "tools/call",
  "params": {
    "tool": "generate_stories",
    "arguments": {
      "project_context": { ... },
      "epics": [/* approved epics from step 1-2 */]
    }
  }
}

Related MCP Services