flow-engineer-subagent

Create IDE-specific subagents with YAML frontmatter and system prompts.

3|Updated Oct 5, 2025
One-click install
npx skills add https://github.com/korchasa/flow --skill flow-engineer-subagent
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: flow-engineer-subagent
Source: https://github.com/korchasa/flow/tree/main/framework/skills/flow-engineer-subagent
Command: npx skills add https://github.com/korchasa/flow --skill flow-engineer-subagent

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Subagents enable developers to define focused AI assistants with isolated context, allowing domain-specific workflows and environment-aware prompts to be created, tested, and reused across IDEs.

Core Features & Use Cases

  • Focused specialization: one task per subagent to maximize reliability and clarity.
  • Cross-IDE placement and detection: supports Cursor, Claude Code, and OpenCode project or user scopes.
  • Frontmatter-driven configuration: YAML frontmatter (name, description, and optional fields) plus a system prompt body for behavior.
  • Phase-based creation workflow: Discovery, Design, Implementation, Verification, and Iteration phases with checklists.
  • Safe testing and iteration: validation steps to ensure proper delegation and guardrails.

Quick Start

Create a new subagent file in the appropriate IDE directory (Cursor, Claude Code, or OpenCode) and populate it with valid YAML frontmatter and a system-prompt body.

Frequently Asked Questions about flow-engineer-subagent

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

FAQPage Schema
How do I create an AI subagent for a specific IDE workflow?

To create an AI subagent, you write a file with YAML frontmatter defining the name and description, followed by a system prompt body. This configures focused task expertise for specific IDE environments.

What is the best way to configure isolated context for domain-specific AI assistants?

Configuring isolated context for domain-specific AI assistants is achieved by creating focused subagents. You define one task per subagent using YAML frontmatter and a structured system prompt to maximize reliability.

Does subagent configuration work across different development environments?

Yes, subagent configuration works across Cursor, Claude Code, and OpenCode environments. The system supports cross-IDE placement and detection for both project-scoped and user-scoped agents.

How do I verify a new subagent is safely configured for deployment?

You verify safe subagent deployment by running Git-ready validation steps during the Verification phase. This ensures proper delegation and validates guardrails before the agent is reused.

What phases are involved in designing a focused AI subagent?

Designing a focused AI subagent involves five phase-based workflows: Discovery, Design, Implementation, Verification, and Iteration. Each phase includes checklists to guide creation and ensure safe testing.