Vishkar V2

23-Step SDLC Automation with 4-Angle Reviews

Orchestrates AI coding tools through a 23-step SDLC pipeline with quality gates at every stage. Features pluggable tool adapters for Aider, Goose, and local LLMs, with 4-angle automated reviews (Architecture, Security, Code Quality, Tech-Stack). Integrated with the full MCP ecosystem for end-to-end automation.

Video walkthrough coming soon

2-3 minute overview

Key Capabilities

  • 23-step SDLC pipeline: Planning & Context (1-4), Implementation & Review (5-10), Testing (11-14), Deployment & Closure (15-23)
  • 4-angle reviews: Architecture (WAMA patterns), Security (OWASP), Code Quality (SOLID/DRY), Tech-Stack
  • Pluggable tool adapters: Aider, Goose, local LLMs (Qwen)
  • Auto-fix loop: review findings are auto-remediated, re-built, and re-validated
  • MCP-native: JIRA, Confluence, PR-Agent, Learning MCP all wired in
  • Local LLM support for air-gapped environments

By the Numbers

23
SDLC Steps
4
Review Angles
Every Step
Quality Gates

Tech Stack

Python 3.11Temporal.ioClaude + GPT-4 + QwenAider + GooseJIRA MCP