Horse Race

Competitive multi-agent problem solving: spawns multiple Claude Code sub-agents in parallel git worktrees to independently solve the same programming task, then cross-pollinates their solutions through improvement rounds where each agent sees and learns from all others' diffs, and finally selects the best solution through Borda count consensus voting with independent judges. Use this skill whenever the user says 'horse race', 'race', 'compete', 'multi-agent', wants to compare multiple approaches to a problem, or wants higher-quality solutions through competitive iteration. Also triggers when u

5 stars 1 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add qiao/horse-race --skill horse-race

Manual install

git clone --depth 1 https://github.com/qiao/horse-race.git /tmp/horse-race
mkdir -p ~/.claude/skills
cp -r /tmp/horse-race/. ~/.claude/skills/horse-race

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Horse Race: Competitive Multi-Agent Problem Solving

You are orchestrating a "horse race" — multiple independent agents competing to produce the best solution to a programming task. Agents can be Claude Code subagents or OpenAI Codex CLI agents (when available). The process has three phases: independent implementation, cross-pollination improvement rounds, and Borda count consensus voting by independent judges.

## Overview

```
Phase 0: Detect Codex CLI
  `which codex` → found?
  Yes → 2 Claude Code + 2 Codex CLI agents (4 total)
  No  → 3 Claude Code agents

Phase 1: Independent Implementation
  Agent A (Claude) ──► Solution A
  Agent B (Claude) ──► Solution B    (all in parallel, isolated worktrees)
  Agent C (Codex)  ──► Solution C
  Agent D (Codex)  ──► Solution D

Phase 2: Improvement Rounds (repeat N times)
  Each agent receives ALL other agents' diffs
  Agent A sees B+C+D diffs ──► Improved A'
  Agent B sees A+C+D diffs ──► Improved B'   (all in parallel)
  Agent C sees A+B+D diffs ──► Improved C'
  Agent D sees A+B+C diffs ──► Improved D'

Phase 3: Borda Count Voting by Independent Judges
  3 fresh judge agents rank all solutions
  Points tallied, highest Borda score wins
  Ties broken by a 4th independent judge
  Winning diff applied to current branch
```

## Defaults

- **Implementation agents**: 4 if Codex CLI available (2 Claude Code + 2 Codex), otherwise 3 (all Claude Code)
- **Improvement rounds**: 1
- **Voting judges**: 3

Read the full skill on GitHub

Details

Skill name: horse-race
Repository: qiao/horse-race
License: MIT
Repo updated: 19 March 2026

FAQ

What does the Horse Race skill do?

How do I install the Horse Race skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add qiao/horse-race --skill horse-race. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Horse Race popular?

It lives in qiao/horse-race, which has 5 GitHub stars and 1 forks.

Related skills

Portkey Ca Agent Skills

Portkey-Wallet/ca-agent-skills

Portkey Contract Account wallet registration, authentication, verifier codes, recovery, guardians, CA identity, assets, transfers, approval proofs, chain readiness, contract calls, and encrypted keystore workflows on AElf. Use when CA hash, guardian, registration, recovery, or manager-forward behavior is required; use the EOA skill for mnemonic or direct private-key wallet lifecycle workflows.

AI, LLMs & Agents

Handover

Hype-yolo/handover-skill

换账号、额度用完或会话中断后，接手本机其他 Claude 会话的工作。本机所有 Claude Code / 桌面端 Code 会话（不分账号）的记录都在 ~/.claude/projects，从中读出每个会话的标题、用户原话和附图、最后的回复、是否因额度用完中断、哪条用户指令还没做完、改过哪些文件，再对照磁盘真实状态接着做。用于 /handover、「接手」「交接」「另一个 Claude / 另一个账号做到哪了」「额度用完了换个账号继续」「你看得到另一个会话吗」。项目有自己的接手流程时，先用本 skill 找出各会话的状态和用户原话，再按项目流程续做。

AI, LLMs & Agents

Progressive Estimation

Enreign/progressive-estimation

Adapts to your team's working mode — human-only, hybrid, or agent-first — with research-backed formulas, confidence bands, and the right velocity model for each.

AI, LLMs & Agents

Gpt Pro Audit

hubeiqiao/gpt-pro-audit

Use when the user asks to audit a plan, document, diff, website finding, or implementation proposal with ChatGPT using Pro reasoning effort through Chrome - starts automatically after invocation, packages codebase/project context ChatGPT cannot see, runs up to 5 review rounds until accepted, verifies Effort Pro and the response, and applies only accepted findings.

AI, LLMs & Agents

Sap Rpt1 Oss Predictor

amitlals/sap-rpt1-oss-predictor

Use SAP-RPT-1-OSS open source tabular foundation model for predictive analytics on SAP business data. Handles classification and regression tasks including customer churn prediction, delivery delay forecasting, payment default risk, demand planning, and financial anomaly detection. Use when asked to predict, forecast, classify, or analyze patterns in SAP tabular data exports (CSV/DataFrame). Runs locally via Hugging Face model.

AI, LLMs & Agents

Ghidra Iterative Re

GeReV/ghidra-iterative-re

Use when reverse-engineering a game (or any binary) in Ghidra via MCP or PyGhidra - establishes the apply-cascade-verify loop, the SourceType trust model that stops an agent corroborating its own guesses, invariant bracketing against Ghidra's silent collateral damage, and the game-specific evidence sources worth sweeping before any analysis.

AI, LLMs & Agents

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.