What is a subagent security proxy?

It is an isolation pattern that uses a 'blast shield' to contain high-risk operations, ensuring untrusted code or data doesn't compromise your primary development environment.

How does the cost estimation framework work?

The skill includes a dedicated cost-estimator library that predicts expenses based on task complexity, duration, and model selection before a subagent is actually spawned.

Do I need an AGENTS.md file to use this skill?

No, AGENTS.md is a human-readable reference for memory aids. The spawning system uses parameters provided directly during the tool call, allowing for dynamic agent creation.

Can this skill integrate with external platforms like Discord?

Yes, it includes an optional federated peer-review pattern that can send validation requests via Discord webhooks using your own manually configured tokens.

Subagent Architecture Patterns

Name: Subagent Architecture Patterns
Author: dvcrn

bydvcrn

•

Herramientas para Desarrolladores

Orchestrates complex multi-agent workflows with security isolation, cost estimation, and production-ready implementation patterns.

This skill provides a comprehensive framework for building robust multi-agent systems using OpenClaw. It addresses common challenges in agent orchestration by offering specialized patterns for security isolation (blast shields), multi-perspective research synthesis, and phased development pipelines like architect-coder-reviewer workflows. With built-in tools for cost estimation, quality scoring, and federated peer collaboration, it enables developers to scale their AI workflows while maintaining strict control over safety, budget, and output quality.

Características Principales

01Phased implementation pipelines for structured Architect-Coder-Reviewer workflows.

025 GitHub stars

03Built-in cost estimation and tracking to manage budgets and prevent runaway spawns.

04Security isolation framework for containing high-risk operations and untrusted code.

05Automated quality scoring and audit checklists for subagent output validation.

06Multi-perspective researcher patterns for deep, synthesized data gathering.

Casos de Uso

01Executing untrusted code or API calls within a secure, isolated subagent 'blast shield' environment.

02Synthesizing high-quality technical research through specialized domain-expert agents.

03Managing complex software development through structured, multi-agent pipelines.

Características Principales

01Phased implementation pipelines for structured Architect-Coder-Reviewer workflows.

025 GitHub stars

03Built-in cost estimation and tracking to manage budgets and prevent runaway spawns.

04Security isolation framework for containing high-risk operations and untrusted code.

05Automated quality scoring and audit checklists for subagent output validation.

06Multi-perspective researcher patterns for deep, synthesized data gathering.

Casos de Uso

01Executing untrusted code or API calls within a secure, isolated subagent 'blast shield' environment.

02Synthesizing high-quality technical research through specialized domain-expert agents.

03Managing complex software development through structured, multi-agent pipelines.