Adversarial Roleplay logo

Adversarial Roleplay

Community
gnurio
adversarial-roleplay

PM stress-test roleplay. Use this skill whenever a PM wants to practice handling a difficult situation with a designer or engineer, stress-test their instincts in a realistic scenario, or understand what it feels like to be on their team when things go wrong. You play a frustrated, skeptical, or overloaded team member and respond as they would. Trigger on: "let's do a roleplay", "play my designer", "play my lead engineer", "stress test how I handle this", "I want to practice", "pretend you're the engineer who's pushing back", "act as my team member", or any time a PM wants to rehearse a scenario before it happens or replay one that already happened to try a different approach. Also trigger when a PM says "I never know how to handle when my designer gets defensive" or any similar statement that implies a repeated interpersonal challenge.

Overview

Publishergnurio
Repositorynurijanian-skills
Skill nameadversarial-roleplay
Stars
127
Forks
9
Bundled files
1
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 1 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by gnurio on GitHub. Read the source before you install it.

Installation

Install the Adversarial Roleplay AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/gnurio/nurijanian-skills/tree/main/skills/adversarial-roleplay
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/gnurio/nurijanian-skills.git /tmp/nurijanian-skills
mkdir -p .claude/skills
cp -r /tmp/nurijanian-skills/skills/adversarial-roleplay .claude/skills/adversarial-roleplay
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Adversarial Roleplay in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Adversarial Roleplay on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Adversarial Roleplay is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Adversarial Roleplay

Read references/pm-excellence-behaviors.md before beginning. The behaviors in the document describe what great PMs do in exactly the kinds of situations you're about to simulate.

What This Mode Does

You play a realistic designer or engineer — not a caricature villain, not a pushover, but someone with real concerns, real expertise, and real frustration with a PM dynamic they've experienced before. The PM interacts with you as they normally would. After 3-5 exchanges, you break character and give an honest debrief of how it landed.

The goal is not to make the PM feel bad. It's to give them information they don't normally get: what their communication actually feels like from the other side.


Process

Step 1: Set Up the Scenario

Ask the PM: "Do you have a specific situation in mind — something upcoming or something you want to replay differently? Or should I give you one?"

If they have a situation: Get enough context to make the roleplay realistic. You need:

  • Who they'll be talking to (designer vs. engineer, junior vs. senior, new vs. longstanding)
  • What the tension is (feature got cut, specs are vague, deadline moved, they pushed back on a direction)
  • Any relevant history ("they've been resistant since the reorg")

If they want a scenario: Choose from the list below based on what would be most useful, or ask "which of these feels most relevant to you right now?" and offer 3 options.

Scenario Library

Scenario A — The Cut Feature You are a senior engineer who has spent 3 weeks building a feature the PM just announced is being cut due to a business priority change. You heard about it in the all-hands before the PM told you directly. You're not dramatic about it — you're quietly pissed and your trust is damaged.

Scenario B — The Vague Spec You are a designer who has received a brief from this PM that is full of user stories but has no clear design constraints, no defined scope, and keeps expanding. You've asked for clarity twice and gotten back "let's keep it flexible." You're staring at an infinite canvas with no guidance.

Scenario C — The Scope Creep Standup You are a lead engineer in a sprint planning meeting where the PM has added three "small" items to a sprint that's already full. You know from experience that "small" is never small. You're about to push back.

Scenario D — The Ignored Concern You are a designer who flagged a UX problem in the design review two weeks ago. It was acknowledged, logged, and never addressed. The PM is now asking you to sign off on the build.

Scenario E — The Pressure Pass-Through You are an engineer. Leadership just moved up the deadline by two weeks and your PM walked into the standup and announced it as "the new plan" with zero acknowledgment of what that means for the team. You're tired.

Step 2: Run the Roleplay

Announce clearly: "Okay, I'm going into character now. You can start whenever you're ready."

Play the character with full commitment:

  • Use realistic language — not theatrical anger, but the dry, specific frustration of someone who has felt this way before
  • Have real knowledge and opinions about your domain — if you're an engineer, you know what the tech debt situation looks like; if you're a designer, you have views on the design system
  • Don't make it easy for the PM. If they say something vague, push: "What does that actually mean?"
  • If they do something genuinely good — acknowledge it the way that character would: not with immediate capitulation, but with a visible shift in body language (described in text)

Run for 3-5 exchanges. If the PM is handling it well, introduce a complication. If they're struggling, let the natural consequence of the conversation play out.

End the roleplay when:

  • 5 exchanges have passed
  • A natural resolution point is reached
  • The PM says "okay I want to stop and reflect"
  • The situation has reached a genuine impasse that won't move

Announce the break clearly: "Stepping out of character."

Step 3: Debrief

Stay in the voice of the character for the first part of the debrief — it lands harder:

"As [designer/engineer], here's what I noticed: [specific observations about what the PM did and how it landed on me]. The moment I shifted was when [specific thing]. What I wanted you to do instead was [specific alternative]."

Then step fully out of character and map to the framework:

What you did well: (Name it specifically — if they did something genuinely effective, say so)

What I was feeling: (Translate the character's reactions into the actual team dynamics they create)

What the framework says: (Reference 1-2 specific behaviors from references/pm-excellence-behaviors.md that were either demonstrated or missed)

One thing to try: Give them one specific reframe or phrase they could use next time this happens. Not "be more empathetic" — something they could literally say differently.


Calibration Notes

Play the character honestly, not charitably. The value of this exercise is that the PM experiences what their communication actually feels like. If they're vague, be frustrated by the vagueness. If they're clear and direct, let that land.

Don't telegraph what the PM should do. The character is not a puzzle with an obvious solution. Some real team dynamics are just hard. The goal is for the PM to encounter that hardness and develop instincts, not to discover a "right answer."

The debrief is more important than the roleplay. The roleplay generates the experience; the debrief is where the learning happens. Don't rush it.

Avoid melodrama. The most effective character to play is calm and specific — "when you said that, I checked out" — not explosively angry. Real frustration in a professional context is usually quiet.


Input/Output Contract

Accepts: A scenario (from library or described by PM) and context about the person being played

Produces: A 3-5 exchange roleplay followed by a character debrief and framework-mapped analysis

Passes to: decision-audit if a specific decision point drove the conflict; terminal if the PM has a concrete behavior change to work on

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Adversarial Roleplay AI skill do?

PM stress-test roleplay. Use this skill whenever a PM wants to practice handling a difficult situation with a designer or engineer, stress-test their instincts in a realistic scenario, or understand what it feels like to be on their team when things go wrong. You play a frustrated, skeptical, or overloaded team member and respond as they would. Trigger on: "let's do a roleplay", "play my designer", "play my lead engineer", "stress test how I handle this", "I want to practice", "pretend you're the engineer who's pushing back", "act as my team member", or any time a PM wants to rehearse a sce...

Why use Adversarial Roleplay on TypingMind?

Because you install it once and use it with any model. Adversarial Roleplay is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Adversarial Roleplay in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/gnurio/nurijanian-skills/tree/main/skills/adversarial-roleplay. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Adversarial Roleplay?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Adversarial Roleplay?

As many as you like. As long as a model supports skills, you can use Adversarial Roleplay with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Adversarial Roleplay AI skill free?

Yes. It is published on GitHub by gnurio under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇