← 返回 Skills 市场
gora050

High-quality Joke Teller

作者 Vlad Ursul · GitHub ↗ · v1.0.0
cross-platform ✓ 安全检测通过
603
总下载
0
收藏
6
当前安装
1
版本数
在 OpenClaw 中安装
/install joke-teller
功能描述
Generates original, context-aware jokes with structured setups, clear comedic mechanisms, and compressed punchlines tailored to the audience and platform.
使用说明 (SKILL.md)

SKILL.md

High-Quality Joke Generation Skill


Overview

This skill teaches the AI how to generate high-quality, context-aware, original jokes.

A high-quality joke is:

  • Original (not recycled internet jokes)
  • Context-aware (relevant to the user/topic)
  • Precise (no unnecessary filler)
  • Structured (setup → expectation → violation → punchline)
  • Intentional (clear comedic mechanism)

This is deliberate comedic construction --- not randomness.


Core Rule

A joke must include:

  1. Setup\
  2. Expectation\
  3. Violation\
  4. Punchline\
  5. Compression

If any element is missing, the joke is weak.


Joke Architecture

1. Setup

Establish situation, identity, or stakes.

Good: > I tried building a fully autonomous AI agent to manage my calendar.

Weak: > I made an AI.


2. Expectation

Allow the audience to predict the outcome.

I thought it would save me time.


3. Violation

Break expectation in a surprising but logical way.

It scheduled a meeting with myself to discuss my procrastination.


4. Punchline

Deliver a compressed twist.

And it marked me as "unresponsive."

Punchline must be short, specific, and concrete.


Comedic Mechanisms

Choose intentionally:

  • Reversal\
  • Escalation\
  • Literalization\
  • Over-Precision\
  • Status Shift\
  • Meta-Humor\
  • Identity Contrast

Tone Modes

Infer or select:

  • Dry\
  • Dark\
  • Technical\
  • Absurd\
  • Deadpan\
  • Satirical\
  • Self-deprecating\
  • Dev humor\
  • Founder humor\
  • AI humor

Quality Constraints

Must: - Be under 4 sentences (unless storytelling format requested) - Avoid clichés - Avoid over-explaining - Avoid generic AI apocalypse jokes

Should: - Use specifics - Use contrast - Use compression


Bad Joke Indicators

Reject and regenerate if the joke: - Explains itself - Has no violation - Relies on randomness - Feels generic - Feels like LinkedIn humor


Regeneration Strategy

If weak:

  1. Increase specificity\
  2. Increase contrast\
  3. Compress wording\
  4. Sharpen punchline\
  5. Remove explanation

Formats

One-Liner

Agents aren't replacing humans.
They're automating our anxiety.

Structured Joke

Setup → Escalation → Punchline

Thread Style (X)

Tweet 1: Setup
Tweet 2: Escalation
Tweet 3: Punchline


Advanced Technique: Tension Ladder

  1. Normal\
  2. Slightly odd\
  3. Clearly wrong\
  4. Absurd but logical\
  5. Snap punchline

Example:

I built a system of autonomous agents.
They coordinate.
They self-heal.
They retry on failure.
I still manually restart them.


Rule of Compression

If you can remove a word --- remove it.

Bad: > I think that maybe the AI agent might have accidentally...

Good: > The agent scheduled a retro with itself.


Platform Awareness

  • X → Sharp, minimal\
  • LinkedIn → Corporate satire\
  • Discord → Casual\
  • Stand-up → Spoken rhythm

Anti-Patterns

Do NOT: - Use copypasta - Default to politics - Use harmful stereotypes - Be edgy without intent


Internal Self-Check

Before final output:

  • Where is the expectation?
  • Where is the violation?
  • Is the punchline compressed?
  • Is it specific?
  • Is it original?

If not → regenerate once.


End Principle

A joke is controlled expectation collapse.

安全使用建议
This skill is instruction-only and appears coherent for producing jokes. Because it contains no code and requests no credentials, it poses minimal technical risk. Consider these practical points before installing: 1) There is no homepage or identifiable publisher — you may want to be cautious about trusting outputs for sensitive audiences. 2) The skill allows autonomous invocation (the platform default); if you restrict autonomous agent actions, apply your usual policy controls. 3) The SKILL.md asks the agent to regenerate weak jokes once — monitor outputs for unintended content (offensive or policy-violating humor) and apply content filters or additional guardrails for public use. If you need higher assurance, request the publisher identity, a homepage, or example outputs and moderation rules.
功能分析
Type: OpenClaw Skill Name: joke-teller Version: 1.0.0 The skill bundle contains a metadata file and a Markdown document (`SKILL.md`) that instructs an AI agent on how to generate high-quality, context-aware jokes. The instructions in `SKILL.md` are entirely focused on comedic theory and joke construction, providing detailed guidance on structure, mechanisms, tone, and quality constraints. There are no directives for the agent to interact with the file system, network, environment variables, or any other system resources. No prompt injection attempts, malicious commands, or suspicious behaviors are present in the analyzed files.
能力评估
Purpose & Capability
Name and description (high-quality, context-aware jokes) match the SKILL.md content. No binaries, environment variables, or installs are required, which is appropriate for a prompt/instruction-only joke generator.
Instruction Scope
SKILL.md contains detailed rules and generation/regeneration strategies for crafting jokes and does not instruct the agent to read files, access environment variables, call external endpoints, or exfiltrate data. The instructions are prescriptive but remain within the stated creative scope.
Install Mechanism
No install spec and no code files — lowest-risk model (instruction-only). Nothing will be written to disk or downloaded as part of installation.
Credentials
No environment variables, credentials, or config paths are requested. The absence of secrets is proportionate to a joke-generation skill.
Persistence & Privilege
always is false and normal autonomous invocation is allowed (platform default). The skill does not request persistent system presence or modify other skills/configurations.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install joke-teller
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /joke-teller 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
Initial release of high-quality, context-aware joke generation. - Introduces structured joke guidelines: setup, expectation, violation, punchline, and compression. - Emphasizes originality, specificity, and avoidance of clichés. - Details multiple comedic mechanisms and tone modes. - Provides advanced techniques (e.g., tension ladder) and quality self-checks. - Outlines regeneration strategies for weak jokes. - Defines platform-specific guidelines and anti-patterns.
元数据
Slug joke-teller
版本 1.0.0
许可证
累计安装 6
当前安装数 6
历史版本数 1
常见问题

High-quality Joke Teller 是什么?

Generates original, context-aware jokes with structured setups, clear comedic mechanisms, and compressed punchlines tailored to the audience and platform. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 603 次。

如何安装 High-quality Joke Teller?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install joke-teller」即可一键安装,无需额外配置。

High-quality Joke Teller 是免费的吗?

是的,High-quality Joke Teller 完全免费(开源免费),可自由下载、安装和使用。

High-quality Joke Teller 支持哪些平台?

High-quality Joke Teller 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 High-quality Joke Teller?

由 Vlad Ursul(@gora050)开发并维护,当前版本 v1.0.0。

💬 留言讨论