pipeline-orchestrator

Orchestrate end-to-end SDLC pipelines across stack-provider plugins and specialist agents.

9|1|Updated Jun 16, 2026

Search Tags:#agent #pipeline #plugins #sdlc #stack #phases #orchestrate

Ready to run

npx skills add https://github.com/AratKruglik/antigravity-sdlc --skill pipeline-orchestrator-aratkruglik

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: pipeline-orchestrator
Source: https://github.com/AratKruglik/antigravity-sdlc/tree/main/plugins/sdlc/skills/pipeline-orchestrator
Command: npx skills add https://github.com/AratKruglik/antigravity-sdlc --skill pipeline-orchestrator-aratkruglik

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The Pipeline Orchestrator coordinates specialist agents to deliver complete features from requirements to PR by orchestrating multi-phase SDLC pipelines across stack-provider plugins.

Core Features & Use Cases

Reads stack.md profiles from installed plugins and auto-discovers the highest-priority match.
Executes a 5-phase pipeline (BA → Dev → QA → Sec → Docs) plus extra stack-defined phases, coordinating across per-phase agents.
Supports per-phase fan-out for aspect-aware phases (development and QA) and a deterministic, replayable workflow.
Stores phase prompts, context, and outputs under docs/plans for traceability and auditability.
Uses local overrides to customize workflows per project while preserving a canonical base profile.

Quick Start

Invoke the /sdlc:start command with a feature description to launch the orchestrator.

Frequently Asked Questions about pipeline-orchestrator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I orchestrate AI agents across SDLC phases from requirements to pull request?▼

You can orchestrate AI agents across SDLC phases by invoking the /sdlc:start command with a feature description to launch an end-to-end pipeline covering BA, Dev, QA, Sec, and Docs.

How does multi-agent fan-out work for development and QA pipelines?▼

Multi-agent fan-out works by constructing per-phase prompts and coordinating specialist agents for aspect-aware phases like development and QA, executing tasks in parallel before generating final artifacts.

Can I customize SDLC pipeline workflows for specific projects?▼

You can customize SDLC pipeline workflows using local overrides that modify phase planning and execution per project while preserving the canonical base stack profile.

How do stack profiles determine which plugins run in an SDLC pipeline?▼

Stack profiles determine plugin execution by reading stack.md files from installed plugins and auto-discovering the highest-priority match during the stack profile resolution step of the workflow.

Does the SDLC orchestrator check dependencies before running pipeline phases?▼

The SDLC orchestrator runs preflight dependency checks as the first step of its deterministic workflow to validate the environment before stack resolution and phase execution begin.

Where are phase prompts and pipeline outputs stored for traceability?▼

Phase prompts, context, and pipeline outputs are stored under the docs/plans directory to ensure full traceability and auditability of the multi-agent SDLC workflow.