God Tibo Imagen

Name: God Tibo Imagen
Author: nomadamas

nomadamas/god-tibo-imagen

10 installs
177 repo stars
Updated June 13, 2026
nomadamas/god-tibo-imagen

Helps with ai & agent building tasks.

About

god-tibo-imagen is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted development.

god-tibo-imagen
AI & Agent Building
AI-coding skill

God Tibo Imagen by the numbers

10 all-time installs (skills.sh)
+1 installs in the week ending Aug 4, 2026 (Skillselion tracking)
Ranked #11,959 of 16,546 AI & Agent Building skills by installs in the Skillselion catalog
Data as of Aug 4, 2026 (Skillselion catalog sync)

npx skills add https://github.com/nomadamas/god-tibo-imagen --skill god-tibo-imagen

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

[![Listed on Skillselion](https://skillselion.com/badge/skills/nomadamas/god-tibo-imagen.svg)](https://skillselion.com/skills/nomadamas/god-tibo-imagen)

Installs	10
repo stars	★ 177
Last updated	June 13, 2026
Repository	nomadamas/god-tibo-imagen ↗

What it does

Helps with ai & agent building tasks.

Files

SKILL.mdMarkdownGitHub ↗

god-tibo-imagen

Generate images from text prompts (and optional reference images) by running the gti CLI.

How to invoke

Use the gti command. It is installed globally via npm install -g god-tibo-imagen and ships with this repo.

Basic generation

gti --prompt "flat blue square icon" --output ./out.png

With reference images

Pass --image <path> one or more times to use existing images as input.

gti --prompt "Make this cat wear a hat" --image ./cat.png --output ./cat-hat.png

gti --prompt "Combine these two styles" --image ./a.png --image ./b.png --output ./combined.png

Supported input formats: png, jpg/jpeg, gif, webp.

Output size

Pass --size <value> to control output dimensions.

gti --prompt "a sunset over mountains" --size 1536x1024 --output ./sunset.png

Allowed values:

auto (model decides)
1024x1024, 2048x2048 (square)
1536x1024, 2048x1152, 3840x2160 (landscape)
1024x1536, 2160x3840 (portrait)

Dry run

Validate auth and print the request without making a network call.

gti --prompt "flat blue square icon" --dry-run

Required arguments

--prompt <text> — required text prompt
--output <path> — output PNG path (required for live runs)

Prerequisites the agent must check

gti is on PATH (npm install -g god-tibo-imagen if missing).
The user has Codex ChatGPT auth at ~/.codex/auth.json with

auth_mode = chatgpt. If it is missing, stop and tell the user — do not try to install Codex or fabricate auth state.

After running

gti saves the PNG to --output and prints a JSON summary that includes savedPath. Report savedPath back to the user.

interface:
  display_name: "god-tibo-imagen"
  short_description: "Generate images via the god-tibo-imagen Python SDK"
  brand_color: "#3B82F6"
  default_prompt: "Generate an image of a sunset over mountains using god-tibo-imagen"

policy:
  allow_implicit_invocation: false

god-tibo-imagen — Agent Skill

Cross-agent Agent Skill that wraps the god-tibo-imagen Python SDK / Node.js CLI for image generation via Codex's private ChatGPT-authenticated backend.

This skill follows the Agent Skills format and is compatible with any coding agent that supports it (Claude Code, Codex, Cursor, OpenCode, Continue, Gemini CLI, etc.).

Layout

skills/god-tibo-imagen/
├── SKILL.md              # Agent-facing skill definition (YAML frontmatter + body)
├── README.md             # This file (installation + usage notes for humans)
├── scripts/
│   └── wrapper.py        # argparse wrapper around the `gti` Python SDK
└── agents/
    └── openai.yaml       # Optional Codex/OpenAI interface descriptor

Prerequisites

Python 3.10+
pip install god-tibo-imagen (the import name is gti)
Existing Codex ChatGPT auth state in ~/.codex/auth.json

(auth_mode = chatgpt)

Installation

Copy or symlink this directory into your agent's skills path.

Vercel `skills` CLI (recommended for any supported agent)

npx skills add NomaDamas/god-tibo-imagen --skill god-tibo-imagen

Manual installation by agent

Agent	Target path
Claude Code	`.claude/skills/god-tibo-imagen/`
Codex	`~/.codex/skills/god-tibo-imagen/` (or project `.agents/skills/`)
OpenCode	`~/.config/opencode/skills/god-tibo-imagen/`
Cursor / Continue / Gemini CLI / Kiro	`.agents/skills/god-tibo-imagen/` (project)

For example, on macOS to install for Claude Code from this repo:

cp -R skills/god-tibo-imagen ~/.claude/skills/

Verifying

python skills/god-tibo-imagen/scripts/wrapper.py \
  --prompt "flat blue square icon" \
  --output ./test.png \
  --dry-run

A successful dry run prints a JSON payload with "mode": "dry-run" and does not perform a live network call.

License

Same as the parent repository.

#!/usr/bin/env python3
"""Agent-skill wrapper for god-tibo-imagen.

A lightweight CLI wrapper around the god-tibo-imagen Python SDK designed
for invocation from any coding agent that supports the Agent Skills format
(Claude Code, Codex, Cursor, OpenCode, Continue, Gemini CLI, etc.) as well
as direct command-line usage.

Example:
    python wrapper.py --prompt "flat blue square" --output ./out.png --dry-run
"""

from __future__ import annotations

import argparse
import json
import sys

from gti.client import Client


def main() -> int:
    parser = argparse.ArgumentParser(
        description="Generate images via the god-tibo-imagen Python SDK."
    )
    parser.add_argument("--prompt", required=True, help="Image generation prompt")
    parser.add_argument("--output", help="Output file path")
    parser.add_argument("--model", help="Model to use (defaults to SDK configuration)")
    parser.add_argument("--dry-run", action="store_true", help="Dry run mode")
    parser.add_argument("--auth-file", help="Path to Codex auth.json")
    parser.add_argument(
        "--installation-id-file", help="Path to Codex installation_id file"
    )
    parser.add_argument(
        "--image",
        action="append",
        help="Input image path (can be used multiple times)",
    )
    parser.add_argument(
        "--debug", action="store_true", help="Enable debug output"
    )
    args = parser.parse_args()

    client_kwargs: dict[str, str] = {}
    if args.auth_file:
        client_kwargs["authFile"] = args.auth_file
    if args.installation_id_file:
        client_kwargs["installationIdFile"] = args.installation_id_file

    client = Client(**client_kwargs)

    gen_kwargs: dict[str, object] = {
        "prompt": args.prompt,
        "dry_run": args.dry_run,
    }
    if args.model:
        gen_kwargs["model"] = args.model
    if args.output:
        gen_kwargs["output_path"] = args.output
    if args.image:
        gen_kwargs["image_paths"] = args.image
    if args.debug:
        gen_kwargs["debug"] = True

    result = client.generate_image(**gen_kwargs)

    output = {
        "mode": result.mode,
        "savedPath": result.saved_path,
        "responseId": result.response_id,
        "sessionId": result.session_id,
        "revisedPrompt": result.revised_prompt,
        "warnings": result.warnings,
    }
    if result.request is not None:
        output["request"] = result.request
    if result.response is not None:
        output["response"] = result.response

    print(json.dumps(output, indent=2))

    return 0


if __name__ == "__main__":
    sys.exit(main())