agentleFS
Sign inStart free
konglingbukongkonglingbukong/gpt-image-studio/AGENTS.md

gpt-image-studio

AGENTS.md1 starsChanged 33 days agoMITView on GitHub

This folder is an image generation tool: the user's own OpenAI API key + the official gpt-image-2.5 API. Full details are in SKILL.md in this folder. If rules.local.md exists (the user's personal preferences), read it first; it overrides this file where they conflict. Reply to the user in their own language. If python isn't found, try python3, py -3, or the full path (on Windows usually %LOCALAPPDATA%\Programs\Python\Python3xx\python.exe; use the full command from rules.local.md if it has one). Common options: --size…

  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 31,774 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

  • In this repositoryOne other file it gives its agents.

The file

We split this AGENTS.md into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If python isn't found, try python3, py -3, or the full path (on Windows usually %LOCALAPPDATA%\Programs\Python\Python3xx\python.exe.

    line 12
  2. Very specific:

    Web workbench: after the user runs start.bat / start.sh, http://127.0.0.1:8765 shows your images live.

    line 14
  3. Very specific:

    Don't change --out-dir, or the web workbench won't see the image.

    line 20

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

gpt-image — instructions for AI agents

This folder is an image generation tool: the user's own OpenAI API key + the official gpt-image-2.5 API. Full details are in SKILL.md in this folder. If rules.local.md exists (the user's personal preferences), read it first; it overrides this file where they conflict. Reply to the user in their own language.

Three ways to use it — pick what your environment supports

  1. MCP tools (if the user registered scripts/mcp_server.py as an MCP server): call generate_image, check_connection, get_spending or sync_balance. On “同步余额” or “sync my balance”, call sync_balance({}), use your authorized signed-in browser to follow the returned steps, and call the tool again with the observation. Report the amount and observation time only after it returns synced. Use get_spending for a local estimate.
  2. Command line (anything that can run a shell command): write the prompt to a UTF-8 text file, then run
    python <this folder>/scripts/gen.py --prompt-file <file> --size "2K 16:9" --quality high --name <short-name>
    If python isn't found, try python3, py -3, or the full path (on Windows usually %LOCALAPPDATA%\Programs\Python\Python3xx\python.exe; use the full command from rules.local.md if it has one). Common options: --size ("2K 9:16", "4K 16:9", 2560x1440), --quality (low / medium / high / xhigh / max), --ref <image> (repeatable), --n <count>, --check (test connection, free), --json (machine-readable output).
  3. Web workbench: after the user runs start.bat / start.sh, http://127.0.0.1:8765 shows your images live.

Rules

  • It costs the user money. With no spec requested, use 2K + high (about $0.06 per image). For 4K, xhigh, max or several images, tell the user the rough cost first.
  • After every run, tell the user what it cost and today's total (both are in the output), and give them the image path or show the image.
  • Don't change --out-dir, or the web workbench won't see the image.
  • On errors, pass the message to the user as-is. Common causes: no key (add it in the web page's Settings), network (set proxy in config.json), invalid size.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free