← 返回 Skills 市场

Comic To Realistic

Name: Comic To Realistic
Author: maweis1981

作者 Patronum · GitHub ↗ · v1.0.3 · MIT-0

cross-platform ⚠ suspicious

145

总下载

当前安装

版本数

在 OpenClaw 中安装

/install comic-to-realistic

功能描述

Convert comic or anime character images into high-quality, realistic portraits while preserving key facial features and expressions.

使用说明 (SKILL.md)

Skills.md

Skill Name

comic-to-realistic

Description

Convert comic, anime, or illustration-style character images into high-quality, realistic human portraits.

This skill leverages AI-powered prompt enhancement and image-to-image generation to transform 2D artistic styles into photorealistic human visuals, while preserving key identity features such as hairstyle, facial structure, and expression.

Service Overview

🌐 Product Website / Console:
https://ai.ngmob.com
🔗 API Base URL:
https://api.ngmob.com

⚠️ Note:

ai.ngmob.com is used for product access, workflow management, and console operations

api.ngmob.com is used strictly for API requests and workflow execution

Use Cases

Anime / comic character → realistic human conversion
AI-generated avatars and profile images
Game character realism enhancement
Visual content creation for marketing and social media

Inputs

Name	Type	Required	Description
Image Input	string (URL)	✅	Publicly accessible image URL (anime/comic style)

Prompt Guidelines

To achieve the best results, include the following in your prompt:

Style Keywords

realistic / photorealistic / ultra realistic
cinematic

Photography Settings

85mm lens
shallow depth of field
studio lighting / soft lighting

Detail Enhancements

detailed skin texture
natural skin tone
high detail face

Preview

Comic Realistic

Example (Recommended)

{
  "Image Input": "https://example.com/anime.jpg"
}

安全使用建议

This skill appears to do what it says (use an external API to convert anime/comic images to realistic portraits) but has an important inconsistency: the manifest expects a Bearer API_KEY while the skill metadata and SKILL.md do not declare any required credentials or tell you where to obtain one. Before installing or using the skill: 1) Ask the publisher for a homepage, privacy policy, and where the API_KEY should come from; verify the ai.ngmob.com endpoint is trustworthy. 2) Understand that submitting image URLs (especially photos of real people) sends data to a third-party service—check retention, sharing, and consent. 3) Only supply an API key created for this service and with minimal permissions; do not reuse high-privilege or long-lived credentials (e.g., cloud provider keys). 4) If you are uncomfortable with an unknown provider, decline installation or request the author to declare required env vars explicitly (API_KEY) and provide a source/terms of service. Additional evidence (author contact, homepage, or documentation) would raise confidence toward benign; absence keeps this suspicious.

功能分析

Package: comic-to-realistic (xpi) Version: 1.0.0 Description: Convert comic or anime images into realistic human portraits using AI The provided files constitute a configuration for an AI-based image transformation service. The manifest.json defines standard API interactions (POST requests and GET polling) with a specific backend (api.ngmob.com) to process images. There is no executable code, shell script, or logic that attempts to access sensitive local files, environment variables, or perform unauthorized data exfiltration.

能力评估

ℹ Purpose & Capability

The skill's name, README, SKILL.md, and manifest all describe an image-to-image workflow against https://api.ngmob.com, which is coherent with a comic→realistic transformation service. However the author/source are unknown and there is no homepage; the manifest points to a third-party API (ai.ngmob.com/api/v1/workflows/...) rather than a well-known vendor, which reduces transparency.

✓ Instruction Scope

SKILL.md instructs the agent to accept a publicly accessible image URL and a prompt, and to call the external API for image-to-image generation. The instructions do not attempt to read local files, shell history, or unrelated environment data. They do reference product and API base URLs and include prompt guidance; scope remains limited to the stated task.

✓ Install Mechanism

There is no install spec and no code files executed by the platform — this is instruction-only. That minimizes install-time risk (nothing is downloaded or written by an installer).

⚠ Credentials

The manifest includes an Authorization header using 'Bearer {{API_KEY}}' and a polling flow that also uses {{API_KEY}}, but the registry metadata and SKILL.md declare no required environment variables or primary credential. This mismatch is an incoherence: calling the remote API legitimately requires an API key, but the skill does not declare that requirement or explain how the key is provided. That gap could lead to unexpected prompts for credentials or silent failures. Also the manifest posts image URLs to a third-party service — users should consider privacy of images uploaded.

✓ Persistence & Privilege

The skill is not always-on and does not request elevated platform privileges. It does not attempt to modify other skills or system configuration. Autonomous invocation is permitted (default) but this is normal and not by itself concerning.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install comic-to-realistic
安装完成后，直接呼叫该 Skill 的名称或使用 /comic-to-realistic 触发
根据 Skill 的参数说明提供必要输入，即可获得结构化输出

版本历史

v1.0.3

- Updated preview section with new example images showcasing both comic and realistic styles. - No changes made to features, functionality, or inputs.

v1.0.2

- Updated input configuration: simplified to a single "Image Input" field (URL); removed separate prompt requirement. - Added a Preview section with a sample image to help users visualize the output. - Updated example usage to match the new input format. - Clarified input table and removed "prompt" parameter.

v1.0.1

- Added service overview section, including product website and clear API base URL distinctions. - Improved and clarified use cases and input descriptions. - Reformatted and streamlined prompt guidelines for clarity. - Updated overall structure for better readability and consistency. - Removed Chinese content; all documentation is now provided in English.

v1.0.0

comic-to-realistic v1.0.0 - Initial release: Convert comic/anime/illustration images into high-quality realistic portraits. - Supports prompt optimization and automatic enhancement for photorealistic output. - Designed for creative use cases including character visualization, social media, and content design. - Best results with high-res, clear, single-character images; supports async processing via API. - Detailed documentation on optimal prompt structure and workflow integration included.

元数据

Slug comic-to-realistic

版本 1.0.3

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 4

常见问题