Automate spec-first SDLC stages with structured YAML configuration.
npx skills add https://github.com/kvsankar/sarathi --skill sarathi
Please help me install this Agent Skill. Skill: sarathi Source: https://github.com/kvsankar/sarathi/tree/main/skills/sarathi Command: npx skills add https://github.com/kvsankar/sarathi --skill sarathi
Sarathi addresses the challenge of managing complex software development lifecycles by providing a structured, automated approach that enhances traceability, resumability, and incorporates human review.
To initiate the 'spec-create' stage using Sarathi, run: sarathi /prompts:spec-create
High-intent search queries and answers about installing and using this skill.
Spec-first software development lifecycle automation is achieved by integrating AI coding agents with multi-stage workflows and structured YAML configuration to manage stages from specification to review. This approach reduces manual effort and ensures traceability across the entire delivery process.
AI coding agent integration supports human review stages by automating artifact creation and then prompting designated experts for sign-off before implementation proceeds. This ensures critical specifications, designs, and plans maintain quality and compliance through mandatory human checkpoints.
The best way to structure an SDLC workflow for spec-first AI coding is to define sequential stages: specification, design, planning, implementation, testing, and review. Each stage requires structured YAML configuration to automate task execution and maintain process resumability.
To start the specification stage for an automated software delivery workflow, execute the specific prompt command to initiate spec creation. This triggers the AI agent to begin drafting the initial project specifications required for subsequent design and planning phases.
A spec-first approach is required to use automated software delivery workflows effectively, as the system assumes the project already follows this methodology. Without pre-existing specification practices, the structured automation and multi-stage YAML configuration cannot properly sequence the lifecycle stages.
Software delivery automation requires structured YAML configuration for each stage to provide the AI coding agents with deterministic execution parameters. This configuration enables traceability, resumability, and consistent state management across the specification, design, planning, implementation, testing, and review phases.
cdrguruInitialize Gemini CLI configuration and auditor persona for the PACK workflow.
kkestellExecute repository implementation plans with code changes, tests, and reviews.
DalimovichCoordinate multi-agent AI reviews on GitHub pull requests.
rik-simformAutomate requirements traceability maintenance across SDLC tools.
