Skip to main content
Alejandro Sánchez

Creator and lead maintainer of QZX

Alejandro Sánchez

Alejandro Sánchez is the creator and lead maintainer of QZX — Quick Zap Exchange. He builds open, predictable tools for AI agents and cross-platform automation, with a focus on explicit results, recoverable operations, and honest compatibility evidence.

QZX — Quick Zap Exchange, created and maintained by Alejandro Sánchez.

Why QZX exists

QZX began after repeated work with agents and automation exposed the cost of switching syntax, interpreting silent output, and verifying the same routine operation differently across operating systems. The project turns those lessons into documented commands, explicit outcomes, and reproducible evidence.

Current areas of work

  • Predictable command interfaces for AI agents
  • QZX Result Contract v1 as an open interoperability proposal
  • Revision-specific MCP interoperability from 2025-06-18 through 2026-07-28
  • Human output and stable JSON from one result
  • Recoverable safety for consequential operations
  • Honest Windows, Linux, and macOS compatibility evidence

Public technical thesis

Agent tools should expose explicit outcomes, evidence, and limits

Alejandro’s work argues that an AI-facing command should not force a model to infer success from silence or scrape unstable terminal prose. QZX Result Contract v1 turns that position into a downloadable schema, validators, runtime enforcement, conformance fixtures, and an open invitation for independent implementations.

His current interoperability work maps that contract to structured-output MCP revisions from 2025-06-18 through 2026-07-28 using outputSchema, structuredContent, effective MCP isError consistency, and receipts that distinguish canonical schema relationships from the weaker SDK-portable structural_core mode. This is a QZX proposal built on the public MCP specification—not an endorsement by MCP and not a claim of industry adoption.

Reviews, counterexamples, independent implementations, conference conversations, podcasts, and bounded enterprise pilots are welcome. Read Alejandro’s updated MCP 2026 guide.

Follow the work, not just the claims

The source, PyPI package, command catalog, safety model, and compatibility limitations are public. Articles and project materials credit collaborators, reviewers, translators, and sponsors when their work is material.

Views expressed by Alejandro are his own and do not imply endorsement by contributors, sponsors, service providers, or organizations mentioned in QZX materials.