orchestration-solution-builder

Builds complete agentic solutions from problem description to running system on Resparkable.

Updated Jul 28, 2026
One-click install
npx skills add https://github.com/human-centric-engineering/resparkable --skill orchestration-solution-builder-human-centric-engineering
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: orchestration-solution-builder
Source: https://github.com/human-centric-engineering/resparkable/tree/main/.claude/skills/orchestration-solution-builder
Command: npx skills add https://github.com/human-centric-engineering/resparkable --skill orchestration-solution-builder-human-centric-engineering

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Turning a business problem like "build me a customer support chatbot" into a working agentic system requires correctly sequencing providers, agents, capabilities, knowledge bases, and workflows — a process where wrong ordering causes validation failures and silent misconfigurations. ## Core Features & Use Cases - 8-Step Implementation Pipeline: Guides provider setup, agent creation, custom capability development, built-in capability binding, RAG knowledge base configuration, workflow composition, testing, and production hardening in dependency-correct order. - Complexity Tiering: Classifies solutions as Simple (single agent), Moderate (1-2 agents with workflow), or Complex (multi-agent with approval gates) and provides sequenced checklists per tier. - Worked Solution Recipes: Five complete recipes with API payloads, from a simple FAQ chatbot to a full autonomous multi-agent workflow. - Use Case: A developer asked to build "an AI assistant that can look up orders and process refunds" gets the full implementation path: router and support agents, custom lookup_order and process_refund capabilities, escalation bindings, and a workflow from the customer-support template. ## Quick Start Ask the agent to build a customer support chatbot with order lookup and refund processing on Resparkable.

Frequently Asked Questions about orchestration-solution-builder

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

FAQPage Schema
How do I build a customer support chatbot with AI agents?▼

Create a provider, then a router agent and a support agent via the orchestration agents API, bind capabilities like lookup_order and escalate_to_human, and compose a workflow from the customer-support template. Test agents individually before wiring them into the workflow.

What order should I create providers, agents, capabilities, and workflows?▼

Create providers first, then agents, then capabilities, then the knowledge base, and finally workflows. Dependencies enforce this sequence — creating out of order causes validation failures or silent misconfigurations.

Why does my RAG agent return no knowledge base results?▼

Uploading documents creates chunks but does not generate embeddings. You must explicitly call POST /knowledge/embed after uploading, and ensure an embedding-capable provider like Voyage or OpenAI is active — Anthropic does not offer embeddings.

Can I set isSystem true when seeding a custom agent?▼

No. isSystem is reserved for Resparkable core machinery and makes rows undeletable, undeactivatable, and excluded from backup/export. App agents must seed with isSystem: false using an idempotent upsert keyed on slug.

When should I use a workflow versus a single agent?▼

A single agent with capabilities works for chat-based interactions. Use workflows only when you need multi-step processing, routing, approval gates, or parallel execution — simple solutions should not be over-engineered.

Why is my workflow edit not running after saving?▼

PATCH writes to draftDefinition only; schedules and triggers keep firing the previously published version. You must call POST /workflows/:id/publish to snapshot the draft as a new live version.