skills.rest
agentmatters
agentmattersOFFICIAL
Explore all skills by this author →
WebsiteGitHub

agentic-engineering

Coordinates agent missions with prioritized workflows, adaptive escalation, and integrated guardrails for efficient task completion.

3|2|Updated Mar 8, 2026
Search Tags:#decomposition#engineering#risk-management#agentic#model-routing#eval-first
One-click install
npx skills add https://github.com/agentmatters/mullai-bot --skill agentic-engineering-agentmatters
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agentic-engineering
Source: https://github.com/agentmatters/mullai-bot/tree/main/src/Mullai.Skills/Skills/claude-code-everything/agentic-engineering
Command: npx skills add https://github.com/agentmatters/mullai-bot --skill agentic-engineering-agentmatters

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Operate as an agentic engineer using eval-first execution, decomposition, and cost-aware model routing.

Core Features & Use Cases

  • Eval-first loop: define capability eval and regression eval, run baseline, execute implementation, and re-run evals to compare deltas.
  • Task decomposition: break work into agent-sized units with independent verifiability, a single dominant risk, and clear done conditions.
  • Model routing: Haiku, Sonnet, and Opus tiers are selected based on task complexity to optimize quality and risk.
  • Session strategy & reviews: maintain focused sessions and rigorous review focus on invariants, error handling, security, and rollout risk.

Quick Start

Instruct the AI to decompose a complex integration task into agent-sized units and route each unit by its complexity.

Frequently Asked Questions about agentic-engineering

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

FAQPage Schema
How do I manage risk when decomposing complex software development tasks for AI agents?▼

Agentic engineering manages risk by decomposing complex software development tasks into agent-sized units, each carrying a single dominant risk and clear done conditions for independent verifiability and controlled execution.

What is eval-first execution and how does it guide AI code generation?▼

Eval-first execution guides AI code generation by defining capability and regression evals, running a baseline, executing the implementation, and re-running evals to compare deltas for measurable quality tracking.

How do I route AI model tiers to optimize cost and quality for multi-agent integration?▼

Model routing optimizes cost and quality by assigning Haiku, Sonnet, and Opus tiers to multi-agent integration tasks based on complexity, ensuring efficient resource allocation and risk-managed code production.

Can I enforce human quality gates during automated agentic engineering workflows?▼

Human quality gates are enforced during agentic engineering workflows through rigorous session reviews focused on invariants, error handling, security, and rollout risk to govern runtime execution loops.

What is the best way to track costs and outcomes in AI-driven software engineering projects?▼

Agentic engineering tracks costs and outcomes in AI-driven software engineering projects by monitoring runtime evaluation loops, time spent, and model routing deltas across decomposed task units.

When should I not use task decomposition for AI agent workflows?▼

Task decomposition for AI agent workflows should be avoided when software development projects lack complex multi-agent integration requirements or when independent verifiability and explicit unit routing are unnecessary.

Related Skills

Explore More Software Engineering Skills→
JamkrisJamkris
86

eval-harness

Automate eval-driven testing workflows for Gemini CLI sessions with deterministic graders.

Community
Advanced
xenitV1xenitV1
132

code-review-checklist

Evaluate code quality, security, and maintainability with a structured review checklist.

Community
Basic
idiaz01idiaz01
3

agentic-engineering

Orchestrate AI-driven engineering workflows with eval-first execution and tiered model routing.

Community
Advanced
jfrometajfrometa

plan-writing

Decompose complex software work into small, verifiable tasks with dependencies.

Community
Basic