agent-architect

Create OpenCode agents and subagents with prompts, tool permissions, and validation scaffolds.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/walcon/opencode-example --skill agent-architect
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-architect
Source: https://github.com/walcon/opencode-example/tree/main/.opencode/skills/agent-architect
Command: npx skills add https://github.com/walcon/opencode-example --skill agent-architect

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill guides the creation of OpenCode agents and subagents, including prompts, tool permissions, and output formats; it helps teams design modular, isolated-context agents to delegate heavy tasks and improve context efficiency.

Core Features & Use Cases

  • Agent templates and scaffolding
  • Subagent isolation and delegation strategies
  • Tool permissions and model configuration guidance
  • Agent patterns and orchestration

Quick Start

Create a new agent: npx tsx scripts/init.ts <name> Validate an agent: npx tsx scripts/validate.ts <path>

Frequently Asked Questions about agent-architect

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

FAQPage Schema
How do I design multi-agent workflows with isolated contexts?

Multi-agent workflows delegate specialized tasks to subagents in isolated contexts to improve efficiency. This Skill provides templates, tool permissions, and prompt configurations to create agents with explicit roles, model settings, and structured outputs that orchestrate complex operations repeatably.

What's the best way to set up subagent delegation and permissions?

Subagent delegation isolates tasks by configuring granular tool permissions, mode specification (primary or subagent), and model parameters for each agent. Use the init.ts script to scaffold agents with role prompts, temperature settings, and validation scaffolds that enforce permission boundaries.

Can I configure agent prompts and tool access separately?

Yes. Agent-architect lets you specify explicit role and expertise prompts, assign granular tool permissions per agent, and set model and temperature parameters independently. Each agent validates against a structured format, ensuring consistent tool access and output requirements across your workflow.

How do I validate agent configurations before deployment?

Run npx tsx scripts/validate.ts <path> to check agent configuration against output format and permission scaffolds. Validation ensures prompts, tool permissions, model settings, and deployment scaffolding meet requirements for repeatable orchestrations.

Do I need OpenCode experience to create agents with this Skill?

The Skill targets agent design and orchestration patterns; it requires understanding workflow architecture and tool delegation but provides scaffolding and templates to guide setup. Start with npx tsx scripts/init.ts <name> to generate a new agent template with required fields.

What happens when subagents operate in isolated contexts?

Isolated subagent contexts prevent cross-contamination of state and tool access, improving context efficiency and security. Each subagent executes with its assigned prompts, permissions, and model configuration independently, returning results to the primary agent without exposing internal state.