skills.rest
tchawla827
tchawla827COMMUNITY
Explore all skills by this author →
GitHub

agentic-engineering

Organize AI-driven implementation tasks into 15-minute units with quality gates.

1|Updated Mar 27, 2026
Search Tags:#ai#agile#quality-assurance#workflow-design#task-management#code-review#model-selection
One-click install
npx skills add https://github.com/tchawla827/GraphForge --skill agentic-engineering-tchawla827
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agentic-engineering
Source: https://github.com/tchawla827/GraphForge/tree/main/.claude/skills/agentic-engineering
Command: npx skills add https://github.com/tchawla827/GraphForge --skill agentic-engineering-tchawla827

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Framework for AI-driven implementation with human quality oversight. It helps teams plan, execute, and review AI-enabled work by breaking tasks into small, verifiable units, defining success criteria upfront, and enforcing discipline around invariants and edge cases.

Core Features & Use Cases

  • 15-Minute Units: Break work into independent, testable tasks that can run in parallel.
  • Task Structure: Provide a clear template with Objective, Acceptance Criteria, Risks, and Completion Signals.
  • Model Selection Strategy: Use Haiku, Sonnet, and Opus to route tasks by complexity and resource needs.
  • Eval-First Execution: Establish baselines (capability, regression) and define explicit completion criteria before work starts.
  • Code Review Focus: Emphasize invariants, edge cases, error handling, security, and cross-file dependencies.
  • Metrics & Tracking: Capture time-to-complete, test coverage, and escalation rationales for continuous improvement.
  • Quality Gates: Checklists to ensure readiness before starting, before review, and before merge.
  • Common Patterns: Provide repeatable patterns for common AI-enabled development workflows.
  • Philosophy Summary: Humans oversee critical decisions while automation handles repeatable, verifiable work.

Quick Start

Outline your first task as a 15-minute unit, specifying the objective, acceptance criteria, and completion signals.

Frequently Asked Questions about agentic-engineering

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

FAQPage Schema
How do I structure AI-assisted development tasks for human quality oversight?▼

Structure AI-assisted development tasks as 15-minute independent units with explicit objectives, acceptance criteria, risks, and completion signals to enable verifiable parallel execution and human quality oversight.

What is the best way to manage AI model selection for coding tasks of varying complexity?▼

Manage AI model selection by routing tasks based on complexity and resource needs, using smaller models like Haiku for simple work and larger models like Sonnet or Opus for complex AI-assisted development tasks.

How do I establish baselines for capability and regression testing before AI code generation?▼

Establish baselines for capability and regression testing before AI code generation by defining explicit completion criteria and evaluating the existing system's behavior to create an eval-first execution checkpoint.

What should a code review focus on when reviewing AI-generated code?▼

Code review for AI-generated code should focus on verifying invariants, edge cases, error handling, security vulnerabilities, and cross-file dependencies to ensure human oversight catches critical systemic issues.

Can I use this workflow design for scaling parallel AI implementation across a software team?▼

Yes, this workflow design scales parallel AI implementation by breaking work into 15-minute testable units, enforcing quality gate checklists before starting and merging, and tracking metrics like time-to-complete and escalation rationales.

When should I not use a 15-minute task unit framework for AI-driven implementation?▼

Avoid using a 15-minute task unit framework for AI-driven implementation when work cannot be broken into independent, verifiable pieces or when tasks lack explicit completion criteria and measurable invariants for quality gates.

Related Skills

Explore More Software Engineering Skills→
rivaldiekaptrrrrivaldiekaptrrr

eval-harness

Define and run eval suites to validate AI task performance.

Community
Advanced
lepahclepahc
1

kspec-merge

Merge approved Git task branches into integration branches with kspec review gates.

Official
Intermediate
andreadellacorteandreadellacorte
5

groove-work-compound

Document lessons learned and update project rules, templates, and documentation.

Community
Advanced
thangvawnthangvawn

agentic-engineering

Automate engineering workflows with AI agents and human oversight.

Community
Advanced