agentic-development

Orchestrate end-to-end software engineering execution across subagents with verification gates.

Updated Feb 13, 2026
One-click install
npx skills add https://github.com/alvarovillalbaa/plugins --skill agentic-development-alvarovillalbaa
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agentic-development
Source: https://github.com/alvarovillalbaa/plugins/tree/main/engineering/skills/agentic-development
Command: npx skills add https://github.com/alvarovillalbaa/plugins --skill agentic-development-alvarovillalbaa

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Orchestrating complex software engineering work in an unfamiliar or large repository—where the agent must orient itself, choose an execution strategy, implement changes safely, and verify outcomes without guesswork.

Core Features & Use Cases

  • Repo orientation & instruction-file routing: rapidly discovers repo shape and loads domain depth from specialized frontend/backend skills.
  • Execution-mode selection: chooses between direct execution, subagents, team-of-agents, and supervised loops for larger tasks.
  • Design-first discipline: enforces brainstorming/design gates for non-trivial work and routes through structured intake.
  • Planning, verification, and review discipline: uses evidence-based gates for implementation, review, and finishing claims.
  • Debugging with causal-chain rigor: routes to investigation workflows to avoid symptom-only fixes.
  • Cross-domain coordination: coordinates frontend/backend changes and multi-agent handoffs with clear roles.

Quick Start

Ask the AI agent: Fix the flaky integration test by finding the root cause and landing a verified patch with a causal chain explanation and updated verification evidence.

Frequently Asked Questions about agentic-development

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

FAQPage Schema
How do I orchestrate multi-agent workflows for complex software engineering tasks in an unfamiliar repository?

Multi-agent workflows orchestrate complex software engineering by choosing an execution model, planning work, and coordinating verification and review across subagents. This approach enforces decision gating and evidence-first completion rules to navigate unfamiliar repositories safely.

What is the best way to debug a flaky integration test using a causal chain investigation?

Causal chain debugging routes to investigation workflows to avoid symptom-only fixes. It enforces evidence-based gates for implementation and finishing claims, ensuring the agent finds the root cause and lands a verified patch with updated verification evidence.

How do I coordinate frontend and backend changes across multiple agents during a PR review?

Cross-domain coordination routes to domain-specific frontend and backend skills to manage multi-agent handoffs with clear roles. It applies structured intake and reference-driven workflows for triage, investigation, and verification across the codebase.

Do I need design-first discipline for non-trivial implementation tasks in large repositories?

Design-first discipline enforces brainstorming and design gates for non-trivial implementation tasks in large repositories. It requires routing through structured intake and planning to ensure safe code modifications and evidence-based review.

Can I use execution-mode selection for orchestrating parallel cloud work and releases?

Execution-mode selection chooses between direct execution, subagents, team-of-agents, and supervised loops for larger tasks. It applies to orchestrating parallel cloud work, releases, and systematic learning loops by routing work appropriately.

What are the limitations of evidence-based gates when doing architecture analysis in complex repos?

Evidence-based gates require explicit verification evidence before completing architecture analysis in complex repos. The limitation is the strict decision gating and routing overhead, which may slow down rapid, exploratory investigations that lack formal verification.