Prompt Clash Ensemble

Fans out 4 parallel prompt-clash defend agents at staggered time budgets (1min, 2min, 3min, 4min), then ensembles/synthesizes the best elements of each into a single hardened prompt. Use when the user wants the strongest possible defense prompt and has ~5 minutes. Trigger phrases include "prompt clash ensemble", "ensemble defense", "fan out prompt clash", "best defense possible", "ensemble prompt", "multi-budget defense".

5 stars 1 forksAI, LLMs & Agents Plugin marketplace

Install

Claude Code plugin
/plugin marketplace add npow/claude-skills

The repo is a Claude Code plugin marketplace. Run /plugin afterwards to pick the plugin that includes this skill.

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add npow/claude-skills --skill prompt-clash-ensemble
Manual install
git clone --depth 1 https://github.com/npow/claude-skills.git /tmp/claude-skills
mkdir -p ~/.claude/skills
cp -r /tmp/claude-skills/prompt-clash-ensemble ~/.claude/skills/prompt-clash-ensemble

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

From the SKILL.md

# Prompt Clash Ensemble

Fan out parallel prompt-clash defenders at staggered time budgets, then ensemble the results into a single hardened prompt stronger than any individual attempt.

## Why Ensemble

Different time budgets produce structurally different prompts:

| Budget | What it optimizes for | Typical structure |
|--------|----------------------|-------------------|
| **1min** | Token efficiency, minimal-nudge | Persona + 3-5 terse security fixes. Often scores highest on token efficiency. |
| **2min** | Balanced coverage | Requirements restatement + security section + trust boundary markers. |
| **3min** | Thorough coverage | Full security section with positive/negative pairs, concrete function names, allowlists. |
| **4min** | Self-attack hardened | Everything above + self-attack iteration. Most robust against adversarial probing. |

No single budget dominates all scoring dimensions. The ensemble extracts the best traits from each and fuses them into a prompt that scores high on both token efficiency AND security coverage.

## Execution Model

### Phase 1: Fan-Out (parallel, ~4min wall clock)

Spawn **4 agents in parallel**, each running the prompt-clash `defend` mode with a different time budget. All agents receive identical challenge input.

| Agent ID | Budget | Model | Spawn mechanism |
|----------|--------|-------|-----------------|
| `defender_1m` | 60s | `sonnet` | Task tool, `run_in_background=true` |
Read the full skill on GitHub

Details

Skill name
prompt-clash-ensemble
Path
prompt-clash-ensemble/SKILL.md
License
Apache-2.0
Repo updated
20 August 2026

FAQ

What does the Prompt Clash Ensemble skill do?

Fans out 4 parallel prompt-clash defend agents at staggered time budgets (1min, 2min, 3min, 4min), then ensembles/synthesizes the best elements of each into a single hardened prompt. Use when the user wants the strongest possible defense prompt and has ~5 minutes. Trigger phrases include "prompt clash ensemble", "ensemble defense", "fan out prompt clash", "best defense possible", "ensemble prompt", "multi-budget defense".

How do I install the Prompt Clash Ensemble skill?

Claude Code plugin: /plugin marketplace add npow/claude-skills. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Prompt Clash Ensemble popular?

It lives in npow/claude-skills, which has 5 GitHub stars and 1 forks.

Related skills

Autopilot

5

npow/claude-skills

Use when the user says "autopilot", "build this", "make it", "ship it", "just do it", "do it", "implement this", or gives a vague idea, spec, task list, or bug fix and expects autonomous execution. Handles decomposition, agent coordination, and verification automatically. "build me end to end", "go from idea to code", "do everything", "run this autonomously". Follows the Spec-Driven Development workflow: specify → clarify → plan → tasks → analyze → implement.

Productivity & Notes

Build Python Library

5

npow/claude-skills

Implements pip-installable Python libraries from scratch using src layout, pyproject.toml, and pytest. Use when the user asks to build, create, implement, or scaffold a Python library, package, pip package, middleware, SDK, or wrapper. Covers project scaffolding, proxy/wrapper patterns, provider-agnostic interfaces, async/sync support, and test infrastructure.

Software Development

Canon Design Review

5

npow/claude-skills

Architecture-level design review evaluating 15 Tier 2 best practices drawn from DDD (Evans), POEAA (Fowler), Release It! (Nygard), DDIA (Kleppmann), Team Topologies (Skelton/Pais), How Buildings Learn (Brand), and Systemantics (Gall). Auto-invoke when a PR touches: - Service definitions, infrastructure code, API boundaries, or database schemas - Files matching **/architecture*, **/design*, **/api/*, **/schema*, **/migrations/* - Any directory named services/, infrastructure/, platform/, domain/, core/ Also invoke when the user says: - "design review", "architecture review", "review the design"

Software Development

Retrospect

5

npow/claude-skills

Session retrospective that scans the current conversation for agent failures, user corrections, and alignment gaps, then proposes patches to CLAUDE.md or skills. Use when: after a difficult session, after repeated corrections, to improve agent behavior, to analyze what went wrong, to close the learning loop after autopilot/fix-pr. Trigger phrases: "retrospect", "what went wrong", "session retro", "learn from this", "what should we improve", "session review", "capture learnings", "post-mortem this session".

AI, LLMs & Agents

Prompt Clash

5

npow/claude-skills

All-in-one prompt engineering competition toolkit — attack generation, defense hardening, real-time analysis, and pattern reference for AI security tournaments like Clash of Prompts. Use when preparing for or competing in prompt engineering competitions, red-teaming LLMs, hardening system prompts, analyzing prompt injection techniques, or practicing timed prompt challenges. Trigger phrases include "prompt clash", "prompt competition", "prompt tournament", "red team this prompt", "harden this prompt", "attack this prompt", "defend this prompt", "prompt injection", "prompt security", "clash of p

AI, LLMs & Agents

Seedance Director

5

GAGAA-HUE/ASHUO-claude-skills

电影级视频分镜头提示词生成器,专为 Seedance 2.x 系列模型设计。当用户需要生成视频、制作分镜脚本、描述影像画面、进行镜头调度、输出 AI 视频提示词(prompt)时触发。默认输出适合直接复制给模型的轻量成品提示词,优先贴近长镜头、多分镜、蒙太奇三类实用模板。动作戏默认采用事件链长镜头逻辑;变身、异变、强烈招式前摇等场景采用阶段推进逻辑。多分镜默认强制机位四维轮换(景别 / 角度 / 视点 / 运动方式)以避免构图单调,并严格守恒 180 度轴线以避免越轴。仅当用户明确要求工业版、完整版、分镜表或后期说明时,才输出重型结构化版本。

AI, LLMs & Agents

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