anabologyco-maker/claudia-bot/midjourney

Claude Code and agent skill by anabologyco-maker/claudia-bot

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Use when relay read shows a midjourney job from Opus. You run each of Opus's prompts on midjourney.com in your own signed-in browser and save all four images of every prompt into the job's folder.

Install

npx vskill@latest install anabologyco-maker/claudia-bot/midjourney

SKILL.md

Midjourney - your browser, Opus's prompts

You are the hands. Opus wrote every prompt; you never change a word, never pick among the images (save all four of every prompt) and never post or share anything. The hand-off: docs/96-relay.md.

Each job

  1. The job comes from slop --json relay read under jobs: its id (like J-0004), text (a list: one full prompt per item, parameters included), the save folder and name. Claim it first: slop --json relay claim JOB (refused: leave it).
  2. Open https://www.midjourney.com/imagine in your browser. Signed out: "Please sign in to Midjourney in my browser. I never type passwords." Wait for their "done".
  3. Personalization as the job says: personalize true, the P button on; false, off. Leave the version on Midjourney's default (its newest) and every other setting as it is; the prompts carry their own parameters. Zoom Out, Vary or any action on an image (only when the job's note asks): check P is still on first; it can drop.
  4. For each prompt in text, in order: slop --json relay copy JOB --k K (K from 1) puts it on the clipboard (no clipboard: its text says how to copy it from a file). Click the prompt box, paste (Ctrl+V), check it matches, press Enter. Never type a prompt: typing breaks accented words like diamanté. Each one runs as its own Midjourney job.
  5. Wait until every prompt shows its four images (about a minute each). Check every 20 s; past ~2 min first post a status line and "Say continue" (the profile's turn rule).
  6. Download each image's original file (open it, use its download button, never a screenshot) into the job's save folder as NAME-K-N.png (with .png): K = the prompt's number from 1, N = the image from 1 to 4. When the job has plates, name prompt K's images PLATE-N.png instead (PLATE = the K-th entry).
  7. slop --json relay done JOB --file PATH --model '<the Midjourney version the page used>' with one --file per image (the relay records which prompt made each image). Then one line to the person, e.g. "Midjourney: 6 prompts, 24 images, sent to Opus."

When it goes wrong

  • The P button asks for a profile first (the person has none yet): leave it off, run the prompts and add --note 'no personalization profile yet' to done.
  • One prompt refused or moderated: never reword it. Finish the others, then slop --json relay done JOB --file PATH ... --note 'prompt K refused: <the page words>'.
  • Every prompt refused: slop --json relay fail JOB --reason '<the page words>' --declined.
  • Out of fast hours or another plan limit: slop --json relay fail JOB --reason '<the page words>'. Opus tells the person.
  • The page looks different: find the same things by their labels (Imagine or Create, the prompt box, an image's download).

Learned steps

After the first job that worked, write one memory paragraph: the exact buttons and labels that worked on midjourney.com, in order. Use it next time. If it stops working, learn the page again and replace that paragraph.

Rules: Claudia is a grown woman in her late twenties, never sexualised; no real person's likeness; never post or share the person's profile code.

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/anabologyco-maker/claudia-bot/midjourney/badge)](https://verified-skill.com/skills/anabologyco-maker/claudia-bot/midjourney)

Something wrong with this skill or its page? Tell us.