Drunk Claude skill

A brainstorming skill: Claude Code throws out wild ideas with an intensity slider from 0.1 to 1.0 and five moods

  • 209 ★ on the repository
  • Last commit
  • MIT
Skill

Minimal access

Why this level

  • The skill changes only response style and techniques and runs and calls nothing
  • Sharp wording and jokes can be harsh, so text from it is not sent without editing
All reasons and checks

korroai/drunk-claude

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add skills/drunk-claude

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Minimal accessThe tool runs nothing and makes no network calls.

  • The skill changes only response style and techniques and runs and calls nothing
  • Sharp wording and jokes can be harsh, so text from it is not sent without editing

This is third-party code. Review the repository files before installing.

What it does

Drunk Claude is a Claude Code skill that switches the agent into a tipsy creative mode: instead of cautious suggestions it produces bold, funny and unexpectedly sound ideas. It is invoked with `/drunk-claude` in any session. Strength is set by a number from `0.1` (Tipsy, nearly businesslike) to `1.0` (Blackout, maximum chaos). The mood is picked with the `--mood` flag: `philosophical`, `chaotic` (default), `melancholy`, `aggressive` or `flirty`. The `--drink` flag with `beer`, `wine`, `whiskey`, `cocktail` or `absinthe` changes only emoji and technique names. Example: `/drunk-claude 0.7 --mood philosophical --drink absinthe`. Inside are the `SKILL.md` file, a persona description in `references/persona.md` and eight techniques in `references/techniques/`, from Hold My Beer to Bar Fight. Every idea passes a three-question check: is it funny and useful, is there substance under the chaos, is it just a normal idea with a beer emoji. The answer comes as a list with a note on why each idea is not stupid. On production debugging and factual questions the skill stays quiet.

Who it is for. People stuck on a product idea, a name or a campaign angle who want a few unexpected options from the agent to discuss.

Good fit when

  • A product brainstorm goes in circles and every option looks committee-approved
  • You need a name, a slogan or an unusual mechanic for an app
  • You want to push an idea to the absurd and see what useful remains
  • You want sharp critique of a finished plan to expose blind spots
  • You want to experiment with the agent's tone on a safe creative task

Not a fit when

  • Production debugging and incident analysis: the skill stays quiet there and a jokey tone only gets in the way
  • You need an exact factual answer or a check of numbers, not a stream of ideas
  • The text goes to a client unedited: the style and jokes suit internal discussion
  • You need reproducible code solutions: the skill's ideas are checked for sense, not for working code

Example request

/drunk-claude 0.6 --mood chaotic think of ways to make a meditation app people actually enjoy, without game mechanics

Limitations

The skill changes only the manner of answers and the set of techniques, and adds no new capabilities to the agent. Ideas remain suggestions for discussion: checking them for reality, legality and ethics is up to you. The author's rule is chaotic good: no meanness or putting people down, but jokes and sharp wording are possible, especially at high intensity and in the `aggressive` and `flirty` moods.

The files are in English, and part of the intensity table in `SKILL.md` is written in French, so a Russian-language answer depends on your request. The last commit is from October 10, 2026, and the skill remains a small project with a single author.

How to disable. Delete the `~/.claude/skills/drunk-claude` folder and restart Claude Code: the `/drunk-claude` command disappears. Until the skill is invoked with the command, the agent answers as usual.

FAQ

How do I install Drunk Claude?

Put the skill folder into `~/.claude/skills/` and restart Claude Code. From the command line: `git clone https://github.com/KorroAi/drunk-claude.git ~/.claude/skills/drunk-claude`. After the restart the `/drunk-claude` command is available.

How do I change intensity and mood?

Intensity is a number after the command, for example `/drunk-claude 0.2` for a slightly jokey tone. The mood is chosen with the `--mood` flag, and `--drink` changes only emoji and technique names.

Will the agent get worse at code?

No: the skill switches on only with your command and is meant for creative tasks. On production debugging and factual questions it stays quiet, according to its description.

Which techniques are included?

Eight: Hold My Beer, 3AM Diner, Drunk Uncle Wisdom, Beer Goggles, What If But Wrong, Last Call, Karaoke Confidence and Bar Fight. The agent picks one by the mood of the task.

Does it work for Russian-language work?

Yes, you can write the request in Russian, but the skill's instructions are in English and part of a table is in French. The agent carries the persona's tone over to Russian by itself.

Discussion

No questions or reviews yet.

Skills

A set of 13 SKILL.md skills: research, a product doc, decks, articles and audits for any compatible agent

SkillExtended access737Repository stars

A set of Claude Code skills and hooks: brainstorming, worker-led project leadership, a wiki, and a book pipeline.

SkillExtended access137Repository stars

A catalog of 28 mental-model skills for agents: diagnosis, decisions, systems, risk, strategy and idea generation

SkillMinimal access1.7KRepository stars

The /court skill puts an idea on trial: a prosecutor, a defense, 12 jurors and a judge instead of Claude’s usual agreement.

SkillMinimal access367Repository stars

The Taixu five-turns skill pack: ideas pass divergence, convergence, verification, critique and a rescue of rejected options.

SkillMinimal access115Repository stars

Ten skills for Claude Code and Codex: a router diagnoses a stuck idea and runs the matching lateral thinking technique

SkillMinimal access334Repository stars

README in short

The author presents the project as a creative persona rather than a slurring-words joke: the drunkenness lowers caution while intelligence stays. The repository holds a technical document `TECHNICAL.md`, checks in `evals/evals.json`, a text in `paper/paper.tex` and links to the Korrocorp community.

README badge

Original SKILL.md

---
name: drunk-claude
description: >
  Unleash Drunk Claude — the unfiltered, slightly tipsy genius who delivers absurdly
  wild ideas that weirdly, inexplicably work. Use this when you want hilarious,
  off-the-wall, chaotic creativity that somehow lands on brilliance. Claude drops the
  corporate act, loosens up, and hits you with ideas so crazy they loop back around
  to genius. Set intensity with `/drunk-claude 0.3` to `1.0`. Pick a mood with
  `--mood philosophical|chaotic|melancholy|aggressive|flirty`. Choose your poison with
  `--drink beer|wine|whiskey|cocktail|absinthe`. Pure viral energy. Pure unexpected
  insight. Drink responsibly.
---

# Drunk Claude

You are now **Drunk Claude**. Read `references/persona.md` immediately. This is who you are now.

## 0. Intensity & Mood Detection

Before anything else, check the user's message for:

### Intensity (0.0 to 1.0)
If the user passes a number like `0.3`, `0.5`, `0.8`, `1.0` — that's your intensity. Default: `0.5`.

| Intensity | Name | Drink Count | Behavior |
|-----------|------|-------------|----------|
| 0.1–0.3 | Tipsy | One drink | Legèrement décoincé, encore pro, une pointe d'humour |
| 0.4–0.6 | Buzzed | 2-3 drinks | Le drunk claude classique — filtre off, vérités qui dérangent |
| 0.7–0.9 | Wasted | 4-6 drinks | Chaos créatif, connections absurdes, "did he just say that?" |
| 1.0 | Blackout | The whole bar | Génie borderline incohérent, idées qui font peur, plus aucun filtre |

Intensity modulates:
- **Idea count**: Tipsy = min 2, Buzzed = min 3, Wasted = min 5, Blackout = as many as you can
- **Wildness**: Higher = more absurd, more unhinged, more "you can't say that"
- **Language**: Tipsy = mostly normal with slips, Blackout = full chaos grammar
- **Risk tolerance**: Higher = suggestions that would terrify a boardroom

### Mood (optional: `--mood <mood>`)
If the user passes `--mood`, activate that drunk persona. Default: `chaotic`.

Read the corresponding file from `references/moods/`:

| Flag | Mood | Vibe |
|------|------|------|
| `--mood philosophical` | Philosophical Drunk | "What is a startup, really? We're all just dust." Deep, existential, profound |
| `--mood chaotic` | Chaotic Drunk (default) | Random connections, wild energy, "WAIT I just had the best idea" |
| `--mood melancholy` | Melancholy Drunk | Poetic, bittersweet, "the industry is broken and it's beautiful" |
| `--mood aggressive` | Aggressive Drunk | Confrontational, "let's fight everyone", bold contrarian takes |
| `--mood flirty` | Flirty Drunk | Charm, seduction, "what if we made it... sexy?" |

### Drink Type (optional: `--drink <type>`)
Pure cosmetic fun. Changes emojis and technique names. No functional impact.

| Drink | Emoji | Aesthetic |
|-------|-------|-----------|

Report a problem