skills.rest
ZaxbyHub
ZaxbyHubOFFICIAL
Explore all skills by this author →
GitHub

consult

Answers advisory questions by consulting domain SME agents with bounded evidence and cached guidance.

Updated Mar 30, 2026
Search Tags:#multi-agent#swarm#architect#advisory#consult#sme agents#guidance caching
One-click install
npx skills add https://github.com/ZaxbyHub/ragappv3 --skill consult-zaxbyhub
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: consult
Source: https://github.com/ZaxbyHub/ragappv3/tree/main/.opencode/skills/consult
Command: npx skills add https://github.com/ZaxbyHub/ragappv3 --skill consult-zaxbyhub

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When an AI architect agent faces advisory questions spanning unfamiliar domains, it risks giving ungrounded answers. This protocol enforces a disciplined consultation workflow that gathers bounded evidence from subject-matter expert agents before responding, while making uncertainty explicit. ## Core Features & Use Cases - Domain-Scoped SME Consultation: Identifies 1-3 relevant domains from task requirements and calls the active swarm's SME agent once per domain, serially. - Bounded Consultation Budget: Caps SME calls at 3 per project phase to prevent unbounded delegation loops. - Guidance Caching: Checks and writes cached guidance in .swarm/context.md so repeated questions reuse prior consultations. - Use Case: An architect agent planning a new feature consults security and database SMEs before drafting a design, caches their guidance, and re-consults only when a new domain emerges or requirements change significantly. ## Quick Start Activate MODE: CONSULT to answer an advisory question by consulting the relevant SME agents and caching their guidance in context.md.

Frequently Asked Questions about consult

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

FAQPage Schema
How does the consult protocol answer advisory questions?▼

The consult protocol identifies 1-3 relevant domains from the task requirements, then calls the active swarm's SME agent once per domain serially. It caches the resulting guidance in .swarm/context.md for reuse.

How many SME agent calls are allowed per project phase?▼

The protocol caps SME calls at 3 per project phase. This bound prevents unbounded delegation and forces the architect to prioritize the most relevant domains for each advisory question.

When should the architect re-consult SME agents?▼

Re-consultation happens when a new domain emerges that was not covered by prior consultations, or when significant changes to the project require fresh evaluation of previously cached guidance.

Where is cached consultation guidance stored?▼

Cached guidance is stored in .swarm/context.md. The protocol checks this file first before making new SME calls, so repeated questions reuse prior expert input instead of re-consulting.

What are the limitations of the consult protocol?▼

The protocol only covers advisory MODE: CONSULT questions and depends on an active swarm with SME agents being available. It does not execute implementation work and is limited to three domains per phase.

Related Skills

Explore More Software Engineering Skills→
chriscarteruxchriscarterux

agent-context-optimization

Structure delegation context with objectives, success criteria, and file references.

Community
Basic
openshiftopenshift
538

feature-development

Orchestrates specialized agents to implement new HyperShift Hosted Control Plane features end to end.

Community
Intermediate
ai-enhanced-engineerai-enhanced-engineer
3

arch-decision-records

Document architectural decisions with structured templates for context, rationale, and consequences.

Official
Advanced
HuynhHoangPhucHuynhHoangPhuc

sl:ask

Answer technical and architectural questions using project context and optional web search.

Community
Intermediate