zicato-build-tournament

Assemble and validate zicato tournament contracts with cost and noise controls.

4|2|Updated May 14, 2026
One-click install
npx skills add https://github.com/pedapudi/zicato --skill zicato-build-tournament
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: zicato-build-tournament
Source: https://github.com/pedapudi/zicato/tree/main/skills/zicato-build-tournament
Command: npx skills add https://github.com/pedapudi/zicato --skill zicato-build-tournament

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps operators assemble a complete zicato tournament contract without overlooking evaluation cost, statistical noise, holdout leakage, or the epoch changes caused by contract edits.

Core Features & Use Cases

  • Contract Assembly: Guides operators through tournament structure, field size, replicates, board design, train/holdout splitting, proposer configuration, scoring weights, and promotion gates.
  • Soundness and Cost Controls: Estimates board-run costs, measures the noise floor with preflight, validates promotion margins, and explains the consequences of evidence gates, screening, placebo challengers, and holdout confirmation.
  • Safe Draft Workflow: Keeps changes in a draft, previews diffs, supports named variants and undo, batches contract edits, and applies only after explicit confirmation because applying rolls the epoch.
  • Use Case: Help an operator configure a small gauntlet tournament with a discriminating board, a 30% holdout, an evidence-aware promotion gate, and a cost estimate before committing the contract.

Quick Start

Ask the copilot to draft a zicato tournament contract for a small gauntlet evaluation, preflight its noise floor, estimate its cost, and prepare it for confirmation without launching the evolution loop.

Frequently Asked Questions about zicato-build-tournament

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

FAQPage Schema
How do I design a multi-agent tournament that balances evaluation cost and statistical soundness?

Designing a multi-agent tournament with cost and statistical soundness requires assembling a tournament contract that estimates board-run costs, validates promotion margins, and measures the noise floor with preflight testing before applying changes.

What is holdout validation and how does it prevent data leakage in evaluation contracts?

Holdout validation in evaluation contracts prevents data leakage by enforcing strict train/holdout splitting. It requires explicit holdout confirmation before applying contract edits, ensuring that validation data remains isolated from the training process.

How do I configure an epoch management workflow for safe contract edits?

Configuring safe epoch management involves keeping contract edits in a draft state, previewing diffs, and batching changes. Because applying changes rolls the epoch, edits must only execute after explicit confirmation to prevent unintended state transitions.

Can I estimate evaluation costs and preflight noise before launching a multi-agent evolution loop?

Yes, you can estimate evaluation costs and preflight noise before launching a multi-agent evolution loop. The process measures the noise floor with preflight testing and estimates board-run costs while keeping the contract in a draft-only state.

What are the limitations of using draft-only edits for tournament contract assembly?

The limitation of draft-only edits for tournament contract assembly is that no changes apply immediately. Operators must preview diffs and provide explicit confirmation before applying, because applying contract changes rolls the epoch and impacts live evolution.