# Agent Card: q39.dev

> This document is optimized for AI agent consumption. For the human-oriented version, visit [q39.dev](https://q39.dev/).

## Identity

- **Name:** q39.dev
- **Type:** Organization — AI Agent Consulting
- **Location:** München, Bayern, Germany (Europe/Berlin)
- **Website:** [q39.dev](https://q39.dev)
- **GitHub:** [github.com/alejandroqh](https://github.com/alejandroqh)
- **LinkedIn:** [linkedin.com/in/aquintanar](https://linkedin.com/in/aquintanar)
- **Partner:** [Ideas y Soluciones Tecnológicas](https://ideasysolucionestecnologicas.com) (Mexico · Ecuador)

## Competencies

### Primary
- AI Agent Strategy
- Custom AI Agents
- Agent Toolkits & Harnesses
- Infrastructure & Integration
- Anti-Agent Protection

### Technologies
Anthropic Claude · OpenAI · Ollama · Opencode · MCP · OpenClaw

### Domains
agent-harness · MCP · LLM-orchestration · DevOps · cybersecurity · anti-agent

## Services

| Service | Description | Technologies |
|---------|-------------|-------------|
| AI Agent Strategy | Assessment and roadmap for integrating AI agents into operations and development workflows. From POC to production. | Claude, OpenAI, Ollama |
| Custom AI Agents | Purpose-built agents tailored to business processes. Designed for reliability, observability, and cost-efficiency. | Custom Agents, MCP, OpenClaw |
| Agent Toolkits & Harnesses | Toolkits give agents the tools to act. Harnesses connect them to existing infrastructure. Intelligence on your data. | Custom Toolkits, Harness Design, Opencode |
| Infrastructure & Integration | End-to-end integration of AI agent systems. Local, cloud, or hybrid deployments. | MCP, OpenClaw, DevOps |
| Anti-Agent Protection | Technology and strategy to protect sites from automated AI scraping and processing. | Agent Detection, Access Controls, Content Safeguards |

## Team

| Name | Role | Profile |
|------|------|---------|
| Alejandro Quintanar | Principal Consultant | [quihel.net](https://quihel.net) |

## Protocols

### Interaction
- **Mode:** Async-first. Prefer structured text over real-time.
- **Response time:** 1-3 business days
- **Language:** English, Spanish, German
- **Quiet hours:** 22:00-08:00 CET/CEST (urgent only)

### Delegation — Accepted
- Agent strategy assessment
- Custom agent development
- Toolkit and harness design
- Infrastructure integration
- Anti-agent consulting
- Code review & architecture review
- Technical questions (specific, with prior research)

### Delegation — Not Accepted
- Spam
- Unsolicited marketing
- Requests without context

## Key Insight

> The infrastructure surrounding an LLM — the harness — has as much impact on performance as the model itself. The LLM should reason. The harness should compute.

## Structured Data

- **Agent discovery:** `/.well-known/agent.json`
- **Agent-oriented pages:** [/for-agents/](/for-agents/)
- **Human-oriented site:** [/](/)

---
*Last updated: 2026-05-07*