Explain AI Registry SDLC workflow routing, phases, and recovery guidance.
npx skills add https://github.com/gleeb/ai-registry --skill sdlc-guide
Please help me install this Agent Skill. Skill: sdlc-guide Source: https://github.com/gleeb/ai-registry/tree/main/systems/sdlc/skills/sdlc-guide Command: npx skills add https://github.com/gleeb/ai-registry --skill sdlc-guide
It explains how the AI Registry SDLC workflow operates, including routing between planning and execution, phase ordering, agent roles, validation gates, checkpoint/resume behavior, and troubleshooting when the SDLC gets stuck or errors.
/sdlc-continue works and how checkpoint state relates to artifacts on disk.Ask the AI: “How do I continue my SDLC workflow after an interruption, and what should I check if validation fails?”
High-intent search queries and answers about installing and using this skill.
The SDLC workflow routes requests via a Coordinator that directs tasks to the Planning Hub or Execution Hub based on lifecycle phase. Each phase handles specific agent roles, validation gates, and orchestration duties to advance the workflow.
To resume an interrupted SDLC workflow, use the /sdlc-continue command. This leverages checkpoint state tied to artifacts on disk to recover from stuck states, crashed agents, or mid-task interruptions without losing progress.
Brownfield re-planning requires following a specific protocol that includes impact analysis and minimum re-dispatch rules. This ensures changes to the existing SDLC workflow are properly routed and validated before execution continues.
SDLC validation fails when artifacts do not meet phase-specific gate requirements. Troubleshooting involves checking the workflow context, verifying agent roles, and following recovery guidance to address the specific stuck state or error.
Agent roles in the SDLC lifecycle include orchestrators and subagents responsible for planning and execution tasks. Defining these responsibilities clarifies what happens next and ensures accurate routing through validation gates.
Brownfield re-planning is needed when modifying an active or existing software lifecycle. It differs from standard execution by requiring impact analysis and minimum re-dispatch rules to manage changes to established artifacts and states.
imonmiDetermines the optimal BMAD workflow, agent, or action based on user questions and context.
Kastalien-ResearchCoordinate COORDINATOR, ARCHITECT, and DEBUGGER agents on Thoughtbox Hub.
LerianStudioCoordinate six-gate development cycles from PM task files with JSON state persistence.
curiositechManage durable WinDAGs state with checkpointed recovery for LLM executions.