pipeline

Orchestrate multi-agent workflows with sequential, branching, and parallel stages.

Updated Feb 20, 2026

Search Tags:#code review #debugging #parallel execution #agent orchestration #branching logic #workflow pipelines

Ready to run

npx skills add https://github.com/ssccio/cc-on-web-base --skill pipeline-ssccio

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: pipeline
Source: https://github.com/ssccio/cc-on-web-base/tree/main/skills/pipeline
Command: npx skills add https://github.com/ssccio/cc-on-web-base --skill pipeline-ssccio

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the friction of coordinating multiple AI agents manually by letting you define an end-to-end agent workflow where each stage receives the right context.

Core Features & Use Cases

Sequential chaining: Pass outputs from an explore stage into architecture and then into implementation.
Branching and conditional routing: Send work to different agents based on complexity or other detected signals.
Parallel-then-merge workflows: Run research and code exploration concurrently, then synthesize results.
Stateful execution and verification: Persist pipeline state, run completion checks, and keep outputs aligned to the original request.
Presets and custom syntax: Use built-in review/debug/research/refactor/security pipelines or create custom chains and branching logic.

Quick Start

Run a feature review pipeline for a change request by typing: /pipeline review "add authentication and role-based access control".

Frequently Asked Questions about pipeline

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

FAQPage Schema

How do I chain multiple agents to execute an end-to-end codebase workflow?▼

Multi-agent orchestration passes structured context between sequential exploration, architecture, and execution stages, enabling parallel execution and branching logic to complete end-to-end codebase workflows.

What is the best way to run parallel research and code exploration then merge results?▼

Using a parallel-then-merge workflow runs research and code exploration concurrently, then synthesizes and merges the structured outputs into a single unified result for downstream stages.

Can I use conditional routing to send tasks to different agents based on complexity?▼

Yes, you can implement conditional branching logic to dynamically route tasks to different agents based on detected complexity signals or other criteria within your pipeline stages.

How do I automate a security audit and refactoring pipeline for my code?▼

You can automate security audits and refactoring pipelines by applying built-in presets or custom syntax to chain agents, executing sequential stages with retries and verification checks before completion.

Does multi-agent workflow orchestration persist pipeline state during execution?▼

Yes, multi-agent workflow orchestration initializes and persists pipeline state during execution, applying error routing, merging parallel outputs, and running verification checks to keep outputs aligned with the original request.

What are the limitations of using custom syntax for agent workflow pipelines?▼

Using custom syntax for agent workflow pipelines requires parsing pipeline definitions and managing stateful execution manually, meaning you must configure error routing, retry logic, and parallel output merging yourself for complex branching scenarios.