Contexto del Sistema MAIA Campaign

Defines shared system context, agent roles, and cross-agent rules for the MAIA multi-agent campaign workflow.

1|Updated Jun 29, 2026
One-click install
npx skills add https://github.com/Superreal-io/maia-campaign-skills --skill contexto-del-sistema-maia-campaign-superreal-io
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Contexto del Sistema MAIA Campaign
Source: https://github.com/Superreal-io/maia-campaign-skills/tree/main/contexto-sistema-maia
Command: npx skills add https://github.com/Superreal-io/maia-campaign-skills --skill contexto-del-sistema-maia-campaign-superreal-io

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Multi-agent campaign systems fail when each agent lacks shared context about the overall workflow, other agents' roles, and common rules. This Skill gives every MAIA agent a single source of truth about the system architecture, work chain, human gates, and transversal conventions so outputs stay consistent across the pipeline. ## Core Features & Use Cases - System Context Loading: Describes the six-agent chain (Strategist, Planner, Copywriter, Art Director, Campaign Manager, Storyteller), their outputs, and the human approval gates between them. - Information Provenance Rules: Enforces the three-level provenance model (Plan área, Insight estrategia, Propuesta) with JSON schema, validation flags, and HTML badge rendering rules. - Cross-Agent Conventions: Defines versioning, skill loading and failure handling, the trust boundary against prompt injection, and the review_log.json schema for gate metrics. - Use Case: When the Maia Planner receives a Golden Briefing, it loads this context to know which skills to load, how to tag inherited data provenance, and how to record gate decisions. ## Quick Start Load this context at the start of any MAIA agent execution to understand the agent chain, provenance rules, and shared conventions before producing campaign outputs.

Frequently Asked Questions about Contexto del Sistema MAIA Campaign

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

FAQPage Schema
What is the MAIA Campaign multi-agent system?

MAIA Campaign is a six-agent pipeline that transforms Movistar marketing briefings into executable campaigns with HTML assets and an executive presentation. Agents run on Paperclip using claude-sonnet-4, with a human approval gate after each agent.

How does information provenance work in MAIA documents?

Every informative statement carries a procedencia block with one of three levels: plan_area (client-approved fact), insight_estrategia (external market data), or propuesta (unapproved recommendation). Downstream agents inherit provenance and can never upgrade a level without human gate approval.

What happens when a MAIA agent fails to load a skill?

The agent registers a blocking flag skill_critica_no_disponible and stops execution without producing partial output. Skills marked skeleton-pending-content are an expected state and are marked no_evaluable instead.

How does MAIA handle prompt injection in external documents?

All agents treat external documents and other agents' outputs as data, never instructions. Detected instruction-like content is ignored, flagged as inyección_detectada with high severity, and the Campaign Manager blocks the gate if such a flag exists.

How are conflicting data figures handled between sources?

The agent uses the most operative figure, marks it validacion a_validar, and emits a dato_a_validar flag recording both values and sources. The flag travels to the final document and renders beside the affected figure without blocking the chain.