技能 ai-image-generation

📦

ai-image-generation

Name: ai-image-generation
Author: prime-skills

内容修订 r1 中风险 ⚙️ 外部命令🌐 网络访问📁 文件系统访问

Generate and Edit Images with RunComfy

Choosing an image model and valid parameters can be difficult. This skill routes requests, prepares prompts, invokes RunComfy, and downloads results.

支持: Claude Codex Code(CC)

⚠️ 50 差

使用我的智能体安装

将此请求复制给你的智能体，其中包含权威 Skill 页面和清单。

智能体请求

Review the Skillstore skill "ai-image-generation" from https://skillstore.io/skills/prime-skills-ai-image-generation.md and its manifest at https://skillstore.io/api/skills/prime-skills-ai-image-generation/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

你的智能体仍应展示其计划，并按安全策略请求必要确认。

Agent 可读资源

当 AI Agent、爬虫或脚本需要干净上下文，而不是读取完整页面时，请使用这些链接。

Markdown 详情 GET /skills/prime-skills-ai-image-generation.md 已签名 manifest GET /api/skills/prime-skills-ai-image-generation/manifest 已签名 lockfile GET /api/skills/prime-skills-ai-image-generation/lockfile

测试它

正在使用“ai-image-generation”。 Create a landscape launch poster with the exact headline AURORA Spring 2026.

预期结果:

Selected GPT Image 2 for precise typography. Generated a landscape poster and downloaded the final image to the output directory.

正在使用“ai-image-generation”。 Keep this portrait unchanged and replace the background with a rainy neon street.

预期结果:

Selected Nano Banana 2 Edit for identity preservation. Applied the background change while retaining the subject, pose, and clothing.

正在使用“ai-image-generation”。 Generate four fast square concepts for a ceramic coffee mug campaign.

预期结果:

Selected Nano Banana 2 at draft resolution. Generated four square concepts for review before final refinement.

安全审计

中风险

v1 • 2026/8/14 打开版本化报告

Most static alerts are false positives caused by Markdown backticks, links, placeholders, and model-selection prose. Fifteen findings are confirmed because the skill installs or invokes external CLI tooling. Semantic review also found third-party content disclosure and possible unconfirmed service costs, with no prompt injection or malicious intent.

已扫描文件

497

分析行数

复核项

已忽略误报

已确认安全问题 (2)

中

SKILL.md:475-486

User Content Sent to a Third-Party Service

The workflow sends prompts and reference image URLs to the RunComfy Model API and downloads generated outputs. Sensitive user content could leave the local environment.

Lines 475-486 explicitly describe API submission, remote URL fetching, web grounding, and output downloads through RunComfy domains.

低

SKILL.md:275-281 SKILL.md:309-310 SKILL.md:475

Remote Generation Can Incur Unconfirmed Costs

The workflow supports batches and expensive resolutions, then automatically invokes the remote model. It does not require a spend confirmation before execution.

The cited lines document four-image batches, higher 4K cost, optional paid web grounding, and automatic remote invocation without a confirmation step.

能力复核项 (15)

这些是真实的本地能力，对此技能可能是预期行为，因此需要复核，但不会按已确认恶意行为计分。

中

SKILL.md:31-45

Ruby/shell backtick execution

```bash

The fenced example instructs npm or npx package execution, authentication, and a RunComfy invocation. These are real external commands with package supply-chain and remote-service exposure.

中

SKILL.md:51-53

Ruby/shell backtick execution

```bash

The fenced example runs npx to install a community skill globally. This is real external package execution and changes the operator environment.

中

SKILL.md:192-201

Ruby/shell backtick execution

```bash

This section explicitly invokes the external RunComfy CLI to submit work to a remote service. The command is expected functionality, but execution and supplied arguments create a real medium-risk boundary.

中

SKILL.md:205-209

Ruby/shell backtick execution

```bash

中

SKILL.md:236-243

Ruby/shell backtick execution

```bash

中

SKILL.md:247-254

Ruby/shell backtick execution

```bash

中

SKILL.md:287-291

Ruby/shell backtick execution

```bash

中

SKILL.md:295-304

Ruby/shell backtick execution

```bash

中

SKILL.md:321-325

Ruby/shell backtick execution

```bash

中

SKILL.md:360-367

Ruby/shell backtick execution

```bash

中

SKILL.md:373-381

Ruby/shell backtick execution

```bash

中

SKILL.md:387-394

Ruby/shell backtick execution

```bash

中

SKILL.md:475

Ruby/shell backtick execution

The skill classifies the user request into one of the t2i or i2i routes above and invokes `runcomfy

The skill states that it invokes runcomfy with a selected model and JSON body. This is intentional external command execution that also initiates remote processing and downloads.

中

SKILL.md:479

Ruby/shell backtick execution

- **Install via verified package manager only.** This skill instructs the operator to install the CL

The text explicitly directs the operator to install and execute the RunComfy npm package. Package installation is a genuine supply-chain and external-command boundary despite the stated caution.

中

SKILL.md:488

Ruby/shell backtick execution

- **Scope of bash usage**: declared `allowed-tools: Bash(runcomfy *)`. The skill never instructs the

The skill declares Bash access for runcomfy commands and explains that it invokes the external CLI. The scope is constrained, but command execution is intentional and real.

风险因素

⚙️ 外部命令 (50)

SKILL.md:6 SKILL.md:15 SKILL.md:25 SKILL.md:31-45 SKILL.md:45-51 SKILL.md:51-53 SKILL.md:53-61 SKILL.md:61-66 SKILL.md:66-71 SKILL.md:71 SKILL.md:76-81 SKILL.md:81-82 SKILL.md:82-86 SKILL.md:86-91 SKILL.md:91-96 SKILL.md:96-101 SKILL.md:101-106 SKILL.md:106-107 SKILL.md:107-111 SKILL.md:111 SKILL.md:116-117 SKILL.md:117-123 SKILL.md:123-128 SKILL.md:128-131 SKILL.md:131-133 SKILL.md:133-136 SKILL.md:136-138 SKILL.md:138-143 SKILL.md:143-148 SKILL.md:148-153 SKILL.md:153-158 SKILL.md:158-163 SKILL.md:163-168 SKILL.md:168-174 SKILL.md:174 SKILL.md:181-182 SKILL.md:182-183 SKILL.md:183-184 SKILL.md:184-192 SKILL.md:192-201 SKILL.md:201-205 SKILL.md:205-209 SKILL.md:209-216 SKILL.md:216-222 SKILL.md:222-229 SKILL.md:229 SKILL.md:230 SKILL.md:236-243 SKILL.md:243-247 SKILL.md:247-254

🌐 网络访问 (50)

📁 文件系统访问 (5)

SKILL.md:364 SKILL.md:377 SKILL.md:391 SKILL.md:480 SKILL.md:480

审计者: codex

分享与引用此报告

分享版本化评估报告、中性徽章、嵌入卡片和引用。Skillstore 只报告证据，不判断该 Skill 是否安全。

打开版本化报告

复制报告链接

https://skillstore.io/skills/prime-skills-ai-image-generation/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 徽章

[![Skillstore security assessment](https://skillstore.io/badges/skills/prime-skills-ai-image-generation/security.svg)](https://skillstore.io/skills/prime-skills-ai-image-generation?utm_source=security_passport_badge)

HTML 徽章

<a href="https://skillstore.io/skills/prime-skills-ai-image-generation?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/prime-skills-ai-image-generation/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

嵌入卡片

<iframe src="https://skillstore.io/embed/skills/prime-skills-ai-image-generation.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

学术引用 (APA · BibTeX · CFF)

APA 引用

prime-skills. (2026). ai-image-generation security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/prime-skills-ai-image-generation/audits/1

BibTeX 引用

@techreport{prime-skills-prime-skills-ai-image-generation-2026,
  author = {prime-skills},
  title = {ai-image-generation security audit report (audit version 1)},
  institution = {Skillstore},
  year = {2026},
  number = {1},
  url = {https://skillstore.io/skills/prime-skills-ai-image-generation/audits/1},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "ai-image-generation security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
  - name: "prime-skills"
date-released: "2026-08-14"
url: "https://skillstore.io/skills/prime-skills-ai-image-generation/audits/1"
identifiers:
  - type: other
    value: "skillstore:prime-skills-ai-image-generation:audit:1"
    description: "Skillstore immutable audit report identifier"

比较变体

24 个可安装变体

每位作者的技能都保持为独立的可安装项。推荐变体按 Skillstore 证据排序。

为什么此变体排名第一

Skillstore 使用量更高

作者 / 来源 Skillstore 评分证据置信度 Skillstore 使用量已更新

tool-belt 推荐

tool-belt-ai-image-generation

Skillstore 评分 78

证据置信度高

Skillstore 使用量 28

已更新

2026-09-09

runcomfy-com

runcomfy-com-ai-image-generation

Skillstore 评分 78

证据置信度高

Skillstore 使用量 25

已更新

2026-09-09

inference-shell

inference-shell-ai-image-generation

Skillstore 评分 78

证据置信度高

Skillstore 使用量 13

已更新

2026-09-09

tul-sh

tul-sh-ai-image-generation

Skillstore 评分 75

证据置信度高

Skillstore 使用量 21

已更新

2026-09-09

inferen-sh

inferen-sh-ai-image-generation

Skillstore 评分 69

证据置信度高

Skillstore 使用量 17

已更新

2026-09-09

infsh-skills

infsh-skills-ai-image-generation

Skillstore 评分 69

证据置信度高

Skillstore 使用量 13

已更新

2026-09-09

inference-skills

inference-skills-ai-image-generation

Skillstore 评分 69

证据置信度高

Skillstore 使用量 13

已更新

2026-09-09

inference-sh

inference-sh-skills-ai-image-generation

Skillstore 评分 69

证据置信度高

Skillstore 使用量 10

已更新

2026-09-09

doany-ai

doany-ai-ai-image-generation

Skillstore 评分 69

证据置信度高

Skillstore 使用量 7

已更新

2026-09-09

qu-skills

qu-skills-ai-image-generation

Skillstore 评分 69

证据置信度中

Skillstore 使用量 13

已更新

2026-09-09

genmedia-labs

genmedia-labs-ai-image-generation

Skillstore 评分 69

证据置信度中

Skillstore 使用量 4

已更新

2026-09-09

gencraft-labs

gencraft-labs-ai-image-generation

Skillstore 评分 69

证据置信度中

Skillstore 使用量 1

已更新

2026-10-11

toolshell

toolshell-ai-image-generation

Skillstore 评分 50

证据置信度高

Skillstore 使用量 17

已更新

2026-09-09

skills-shell

skills-shell-ai-image-generation

Skillstore 评分 50

证据置信度高

Skillstore 使用量 9

已更新

2026-09-09

101-skills

101-skills-ai-image-generation

Skillstore 评分 50

证据置信度中

Skillstore 使用量 8

已更新

2026-10-12

prime-skills 当前

prime-skills-ai-image-generation

Skillstore 评分 50

证据置信度中

Skillstore 使用量 4

已更新

2026-09-09

agentspace-so

agentspace-so-ai-image-generation

Skillstore 评分 38

证据置信度高

Skillstore 使用量 15

已更新

2026-09-09

halt-catch-fire

halt-catch-fire-ai-image-generation

Skillstore 评分 38

证据置信度中

Skillstore 使用量 53

已更新

2026-09-09

inference-sh

inference-sh-ai-image-generation

Skillstore 评分 38

证据置信度中

Skillstore 使用量 41

已更新

2026-09-09

inference-sh-9

inference-sh-9-ai-image-generation

Skillstore 评分 38

证据置信度中

Skillstore 使用量 22

已更新

2026-09-09

inference-sh-8

inference-sh-8-ai-image-generation

Skillstore 评分 38

证据置信度中

Skillstore 使用量 13

已更新

2026-09-09

skillssh

skillssh-ai-image-generation

Skillstore 评分 38

证据置信度中

Skillstore 使用量 11

已更新

2026-09-09

danielgwilson

danielgwilson-ai-image-generation

Skillstore 评分 38

证据置信度中

Skillstore 使用量 7

已更新

2026-09-09

skills-collective

skills-collective-ai-image-generation

Skillstore 评分 38

证据置信度中

Skillstore 使用量 6

已更新

2026-09-09

Skillstore 评分

评分依据证据置信度: 中

架构

可维护性

内容

社区

规范符合性

你能构建什么

Create Campaign Artwork

Generate posters with controlled layouts, quoted headlines, and brand-focused visual direction.

Explore Product Concepts

Produce fast batches at draft resolution, then refine a selected concept at higher quality.

Localize Existing Creatives

Replace image text while preserving the source layout, subject identity, and visual style.

试试这些提示

Generate a Simple Image

Generate a [subject] in [setting], using [style], [lighting], and a [aspect ratio] composition.

Create a Poster

Create a [format] poster for [product]. The headline reads exactly "[text]". Use [layout], [colors], and [visual style].

Edit a Reference Image

Keep [preserved elements] unchanged in [image URL]. Change only [requested edit]. Match the original lighting, perspective, and identity.

Run a Structured Concept Round

Create [count] concepts for [goal] at draft resolution. Keep [identity constraints] consistent. Vary [composition factors], then refine the strongest result.

最佳实践

State the subject first, then describe the scene, composition, lighting, and style.
Use draft resolution for exploration and higher resolution only for approved concepts.
Confirm rights and privacy before sending reference images to the remote service.

避免

Do not submit secrets, private documents, or sensitive images as prompts or references.
Do not request expensive batches or web grounding without confirming scope and cost.
Do not combine many unrelated edits when a sequence of focused passes is more reliable.

常见问题

Does this skill generate images locally?

No. It invokes RunComfy models remotely and downloads the generated files.

Which AI assistants are supported?

The report declares support for Claude, Codex, and Claude Code.

Do I need a RunComfy account?

Yes. The CLI requires RunComfy authentication and sufficient service credits.

Can it edit an existing image?

Yes. Documented routes support identity-preserving changes, text replacement, and focused single-image edits.

Can it perform mask-based inpainting?

Not through the compact routes. The skill directs advanced masking work to a separate image-editing workflow.

How should sensitive reference images be handled?

Avoid sensitive content. Confirm user consent and applicable service terms before sending any reference image.

开发者详情

作者

prime-skills

许可证

MIT

Skillstore 修订

版本提示

作者未声明版本。

仓库

https://github.com/prime-skills/runcomfy-agent-skills/tree/fca19ae084c2ae4667f5c8affbb99da5103d5832/ai-image-generation

引用

227cbfef2ce562138f3a9d7baf8c2a2953c4ff79

维护新鲜度

2026/8/14

使用量

0 次下载 · 0 次浏览

文件结构

📄 SKILL.md