kalash33/kstack/agent-design
Claude Code and agent skill by kalash33/kstack
Trust tier 3
- community
Source check pending
Checking that the SKILL.md file is still in the repository.
Design multi-agent architectures. Covers agent roles, communication patterns, orchestration strategies, memory systems, and tool integration. Use when: "design an agent", "multi-agent system", "agent architecture", "how should my agents communicate", or planning any AI agent project.
Install
npx vskill@latest install kalash33/kstack/agent-designSKILL.md
The same SKILL.md was published earlier as kalash33/kstack/agent-design.
/agent-design — Agent Architect
Design multi-agent systems that actually work in production.
When to Use
- Starting a new agent project
- Designing multi-agent coordination
- Choosing between frameworks (CrewAI, LangGraph, PydanticAI, custom)
- Planning agent memory and tool integration
Process
Step 1: Define the Problem
- What task(s) need automation?
- What decisions require judgment vs. are mechanical?
- What's the failure mode? (wrong answer vs. no answer vs. slow answer)
- What's the human-in-the-loop boundary?
Step 2: Agent Decomposition
For each capability needed:
- Role: What is this agent responsible for?
- Tools: What can it access? (APIs, databases, file system, browser)
- Memory: What does it need to remember? (short-term context, long-term knowledge)
- Judgment: What decisions does it make autonomously?
Step 3: Choose Architecture Pattern
| Pattern | When to use | |
|---|---|---|
| Single agent + tools | One task, multiple capabilities | |
| Pipeline | Sequential processing (A → B → C) | |
| Supervisor | One coordinator dispatching to specialists | |
| Swarm | Peer agents with handoff protocols | |
| Hierarchical | Manager → team leads → workers |
Step 4: Communication Design
- Handoff protocol: How do agents pass work?
- State sharing: What's shared vs. private?
- Error handling: What happens when an agent fails?
- Human escalation: When does a human get pulled in?
Step 5: Output
Chaining
- After:
/agent-build(implements the design),/agent-eval(tests it)
Security
- Security scanPassed, 95 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/agent-design/SKILL.md
Popularity
- 7-day trend score
- 5
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/kalash33/kstack/agent-design)