pipeline-sdlc/skill/SKILL.md

name	description
pipeline-sdlc	SDLC module — project management, iterations, bug tracking, test management, and AI cockpit.

pipeline_sdlc

Complete SDLC lifecycle management with AI-powered development cockpit.

Architecture

pipeline_sdlc/
├── pipeline_sdlc/
│   ├── init.py              # load_pipeline_sdlc()
│   ├── adapter.py           # LLM context + model management
│   └── handlers/
│       ├── design.py        # Design phase agent
│       ├── develop.py       # Coding phase agent
│       ├── test.py          # Testing phase agent
│       ├── deploy.py        # Deployment phase agent
│       └── ops.py           # Operations phase agent
├── wwwroot/
│   ├── api/
│   │   ├── cockpit_agent.dspy   # Agent task execution
│   │   ├── cockpit_chat.dspy    # Chat + auto-execute
│   │   ├── cockpit_cards.dspy   # Dashboard stats
│   │   └── bug_*.dspy           # Bug operations
│   └── sd_cockpit/index.ui      # Cockpit main UI

Key DSPY Endpoints

Cockpit (AI Development)

• cockpit_agent.dspy — Agent-driven task execution
• cockpit_chat.dspy — Chat with auto-execute
• cockpit_cards.dspy — Dashboard statistics

Bug Management

• bug_confirm.dspy — Confirm bug
• bug_close.dspy — Close bug
• check_iteration_bugs.dspy — Pre-close bug check

Pitfalls

• Cockpit LLM calls: Uses pipeline_service.llm_bridge.llm_call() via ServerEnv.
• Workspace isolation: Each project has workspace_dir — agent file ops scoped here.
• Deliverable storage: Agent outputs saved to pipeline_deliverables with quality scores.
• CRUD generated dirs are auto-generated by xls2ui — never edit directly.
• index_cockpit.ui is the new cockpit entry — ensure RBAC registered.