skills.rest
Dinxeshh
DinxeshhCOMMUNITY
Explore all skills by this author →
GitHub

Nexus

Coordinate specialist agents to execute minimal viable chains with guardrails.

Updated Mar 18, 2026
Search Tags:#routing#multi-agent#orchestration#guardrails#task-decomposition#autorun
One-click install
npx skills add https://github.com/Dinxeshh/agent-skills --skill nexus-dinxeshh
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: Nexus
Source: https://github.com/Dinxeshh/agent-skills/tree/main/skills/nexus
Command: npx skills add https://github.com/Dinxeshh/agent-skills --skill nexus-dinxeshh

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Nexus eliminates the burden of manually managing multi-domain work by decomposing requests, routing them through specialist AI agents, and executing the minimum viable chain in AUTORUN modes with rigorous guardrails and hub-spoke ownership.

Core Features & Use Cases

  • Centralized Routing and Execution: Classify tasks, consult the routing matrix, and select the leanest agent chain while enforcing L1-L4 guardrails and structured handoffs.
  • Operational Modes and Guardrails: Operate in AUTORUN_FULL, AUTORUN, Guided, or Interactive modes with conflict resolution, error recovery chains, and automatic verification before delivery.
  • Proactive and Parallel Orchestration: Scan project state in neutral mode, recommend next work, coordinate parallel branches with conflict detection, and aggregate outputs into the required Japanese delivery format.
  • Use Case: For a complex project launch spanning UI, backend, and security, Nexus breaks down the scope, routes work to Muse, Builder, Sentinel, and Radar as needed, executes in AUTORUN_FULL, and provides a structured ## Nexus Execution Report.

Quick Start

Invoke Nexus with /Nexus your complex task to have it design, route, and run the smallest guarded chain automatically in AUTORUN_FULL.

Frequently Asked Questions about Nexus

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

FAQPage Schema
How does multi-agent orchestration handle task decomposition for complex software requests?▼

Multi-agent orchestration decomposes complex software requests by routing classified tasks through specialist agents, executing the minimum viable chain with L1-L4 guardrails, structured handoffs, and automatic verification before final delivery.

Can I run parallel agent branches with conflict detection during multi-agent routing?▼

Yes, parallel agent branches with conflict detection are supported. The orchestration hub scans project state in neutral mode, coordinates parallel routing matrices, detects conflicts, and aggregates outputs into a structured execution report.

What operational modes are available for guarding automated agent chains?▼

Operational modes include AUTORUN_FULL, AUTORUN, Guided, and Interactive. Each mode enforces L1-L4 safety checks, conflict resolution, error recovery chains, and automatic verification before delivering the final aggregated output.

Do I need specific specialist agents to use routing matrices for multi-domain projects?▼

Specialist agents such as Muse, Builder, Sentinel, and Radar are coordinated by the central routing hub. The system classifies your multi-domain tasks and selects the leanest agent chain to execute the work automatically.

When should I avoid using automated task decomposition for multi-agent execution?▼

Automated task decomposition and AUTORUN modes should be avoided when your project requires manual conflict resolution or lacks clearly defined multi-domain boundaries, as the system relies on structured handoffs and automatic verification.

Related Skills

Explore More Software Engineering Skills→
GeargrindadminGeargrindadmin

intelligent-routing

Analyze user requests to route tasks to specialist agents.

Community
Intermediate
ucirelloucirello
2

project-completion-verification

Scan GOAL.md to count completed and pending project tasks.

Community
Intermediate
PetriLahdelmaPetriLahdelma
2

dispatching-parallel-agents

Dispatch parallel agents to investigate independent task domains concurrently.

Community
Basic
eyadsibaieyadsibai
7

crewai-agents

Create and execute multi-agent systems with CrewAI for task orchestration.

Community
Advanced