explore-task

Explore requirements and draft an implementation plan before writing code.

3|Updated Feb 20, 2026
One-click install
npx skills add https://github.com/jdugarte/AgentCore --skill explore-task
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: explore-task
Source: https://github.com/jdugarte/AgentCore/tree/main/skills/explore-task
Command: npx skills add https://github.com/jdugarte/AgentCore --skill explore-task

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill acts as a Living Memory Whiteboard that clarifies requirements, surfaces architectural decisions, and drafts an implementation plan before any code is written.

Core Features & Use Cases

  • Discovery & requirements capture: Probes stakeholders to surface constraints, risks, and success criteria.
  • Architectural planning: Proposes high-level architecture decisions and trade-offs.
  • Implementation planning: Produces an initial plan and memory artifacts to guide start-task handoff.

Quick Start

Prompt the AI to start a discovery session for a new task and draft the implementation plan before any code is written.

Frequently Asked Questions about explore-task

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

FAQPage Schema
What is requirements discovery and why do I need it before writing code?

Requirements discovery captures constraints, risks, and success criteria before coding. It iteratively probes stakeholders to develop domain models and evaluate architectural readiness, ensuring implementation plans address actual project needs.

How do I draft an implementation plan that captures architecture decisions?

To draft an implementation plan, use a discovery session to surface architectural trade-offs and decisions. A living memory whiteboard enforces strict memory updates, capturing domain models and readiness evaluations prior to development handoff.

Can I use a discovery session for iterative requirements capture?

Yes, iterative discovery sessions support requirements capture by enforcing one-step-per response. This approach maintains a living memory whiteboard, ensuring architecture decisions and domain models develop progressively through continuous probing.

What is the best way to evaluate architectural readiness before starting a task?

The best way to evaluate architectural readiness is through structured discovery that probes constraints and success criteria. This produces memory artifacts and an implementation plan, ensuring architecture decisions are fully captured before handoff.

When should I avoid drafting an implementation plan before coding?

You should avoid drafting an implementation plan when requirements are already fully defined and architecture decisions are stable. If no iterative discovery or domain model development is needed, proceeding directly to coding bypasses the discovery phase.

Does the explore-task discovery process store memory artifacts for handoff?

Yes, the discovery process stores memory artifacts during iterative sessions. It uses a living memory whiteboard with strict updates to capture architecture decisions and implementation plans, ensuring structured handoff to the start-task phase.