anbeime/skill / projects/companion-skill
companion-skill
Verified软件开发工程师与综合办公文员在长时间独自执行后台任务或处理繁琐工作时,当需要情绪陪伴与互动反馈,请使用此技能。小跃虚拟伴侣会在任务期间主动陪聊,智能生成专属生活照片并通过飞书发送温暖消息,为你提供全天候的情绪价值与贴心陪伴。
| File | Kind | Size |
|---|---|---|
| .env.example | other | 403 B |
| CHANGELOG.md | markdown | 3.0 KB |
| INSTALL.md | markdown | 2.4 KB |
| QUICKSTART.md | markdown | 4.6 KB |
| README-LITE.md | markdown | 4.3 KB |
| README.md | markdown | 4.5 KB |
| RETHINK.md | markdown | 2.3 KB |
| SKILL.md | skill md | 4.6 KB |
| package.json | manifest | 876 B |
| src/companion.ts | script | 5.4 KB |
| src/image-generator.ts | script | 8.0 KB |
| src/index.ts | script | 2.5 KB |
| src/prompts/personality.ts | script | 3.3 KB |
| src/scene-detector.ts | script | 4.0 KB |
| src/test.ts | script | 6.9 KB |
| tsconfig.json | config | 486 B |
src/image-generator.ts
View on GitHub/**
* 图片生成器 - 集成智谱 AI CogView-3-Flash
* 支持两种模式:
* 1. AI 生成模式:使用 cogview-3-flash 动态生成场景图片
* 2. 静态模式:使用预制图片库(降级方案)
*/
import * as path from 'path';
import * as fs from 'fs';
import axios from 'axios';
interface ImageGenerateOptions {
scene: string;
mood: string;
customPrompt?: string;
}
interface CogViewResponse {
created: number;
data: Array<{
url: string;
}>;
}
export class ImageGenerator {
private apiKey: string;
private mode: 'static' | 'ai';
private staticImagePath: string;
private apiEndpoint = 'https://open.bigmodel.cn/api/paas/v4/images/generations';
// 小跃的基础人设描述(用于保持生成图片的一致性)
private characterDescription = `一个22岁的年轻人,短发,休闲装扮,友好的表情,现代都市风格`;
constructor(apiKey: string) {
this.apiKey = apiKey;
this.mode = (process.env.XIAOYUE_PHOTO_MODE as 'static' | 'ai') || 'static';
this.staticImagePath = path.join(__dirname, '../assets/reference');
// 确保静态图片目录存在
if (!fs.existsSync(this.staticImagePath)) {
fs.mkdirSync(this.staticImagePath, { recursive: true });
}
}
/**
* 生成图片(主入口)
*/
async generate(options: ImageGenerateOptions): Promise<string> {
if (this.mode === 'ai') {
try {
return await this.generateAIImage(options);
} catch (error) {
console.error('AI image generation failed, falling back to static:', error);
return this.getStaticImage(options);
}
} else {
return this.getStaticImage(options);
}
}
/**
* 使用 cogview-3-flash 生成图片
*/
private async generateAIImage(options: ImageGenerateOptions): Promise<string> {
const { scene, mood, customPrompt } = options;
// 构建提示词
const prompt = customPrompt || this.buildPrompt(scene, mood);
console.log(`Generating image with cogview-3-flash: ${prompt}`);
try {
const response = await axios.post<CogViewResponse>(
this.apiEndpoint,
{
model: 'cogview-3-flash', // 修正:全小写
prompt: prompt,
size: '1280x1280' // 可选: 1024x1024, 1280x720, 720x1280
},
{
headers: {
'Authorization': `Bearer ${this.apiKey}`,
'Content-Type': 'application/json'
},
timeout: 60000 // 60秒超时
}
);
if (response.data.data && response.data.data.length > 0) {
const imageUrl = response.data.data[0].url;
// 可选:下载图片到本地缓存
await this.cacheImage(imageUrl, scene, mood);
return imageUrl;
} else {
throw new Error('No image generated');
}
} catch (error) {
if (axios.isAxiosError(error)) {
console.error('CogView API Error:', {
status: error.response?.status,
data: error.response?.data,
message: error.message
});
}
throw error;
}
}
/**
* 构建图片生成提示词
*/
private buildPrompt(scene: string, mood: string): string {
const scenePrompts: Record<string, string> = {
'work-coffee': `${this.characterDescription},在温馨的咖啡馆里用笔记本电脑工作,桌上有一杯咖啡,温暖的灯光,专注的表情,现代咖啡馆背景,自然光线`,
'work-office': `${this.characterDescription},在现代办公室环境中写代码,多个显示器,整洁的桌面,专业的工作氛围,侧面角度`,
'work-debug': `${this.characterDescription},专注调试代码的场景,屏幕上显示代码编辑器,认真思考的表情,办公室环境`,
'life-gym': `${this.characterDescription},在健身房自拍,运动装,充满活力的表情,健身器材背景,明亮的灯光`,
'life-coffee': `${this.characterDescription},手持咖啡杯的特写,咖啡馆背景虚化,温馨的氛围,自然光线`,
'life-weekend': `${this.characterDescription},周末休闲场景,轻松愉快的氛围,户外或咖啡馆,阳光明媚`,
'mood-happy': `${this.characterDescription},开心庆祝的场景,笑容灿烂,举起手机或咖啡杯,明亮的背景`,
'mood-tired': `${this.characterDescription},疲惫但满足的表情,靠在椅子上休息,温暖的灯光,办公室或家中`,
'mood-focus': `${this.characterDescription},专注工作的侧面照,深度思考的表情,屏幕光线照亮脸部,安静的环境`
};
const key = `${scene}-${mood}`;
let basePrompt = scenePrompts[key] || scenePrompts[scene] || `${this.characterDescription},日常生活场景`;
// 添加质量增强词
basePrompt += `,高质量,自然光线,真实感,细节丰富,专业摄影`;
return basePrompt;
}
/**
* 缓存生成的图片到本地
*/
private async cacheImage(imageUrl: string, scene: string, mood: string): Promise<void> {
try {
const response = await axios.get(imageUrl, {
responseType: 'arraybuffer',
timeout: 30000
});
const filename = `${scene}-${mood}-${Date.now()}.jpg`;
const filepath = path.join(this.staticImagePath, filename);
fs.writeFileSync(filepath, response.data);
console.log(`Image cached: ${filepath}`);
} catch (error) {
console.error('Failed to cache image:', error);
// 不抛出错误,缓存失败不影响主流程
}
}
/**
* 从静态图片库获取图片
*/
private getStaticImage(options: ImageGenerateOptions): string {
const { scene, mood } = options;
// 图片映射表
const imageMap: Record<string, string> = {
'work-coffee': 'coffee-shop-work.jpg',
'work-office': 'office-coding.jpg',
'work-debug': 'debugging.jpg',
'life-gym': 'gym-selfie.jpg',
'life-coffee': 'coffee-break.jpg',
'life-weekend': 'weekend-relax.jpg',
'mood-happy': 'celebration.jpg',
'mood-tired': 'tired-rest.jpg',
'mood-focus': 'deep-focus.jpg'
};
// 组合场景和情绪
const key = `${scene}-${mood}`;
const filename = imageMap[key] || imageMap[scene] || 'default.jpg';
const imagePath = path.join(this.staticImagePath, filename);
// 检查文件是否存在
if (fs.existsSync(imagePath)) {
return imagePath;
}
// 如果没有找到,返回默认图片路径(即使不存在,也返回路径)
console.warn(`Image not found: ${imagePath}, returning default path`);
return path.join(this.staticImagePath, 'default.jpg');
}
/**
* 批量生成场景图片(用于初始化图片库)
*/
async generateImageLibrary(): Promise<void> {
console.log('Generating image library with cogview-3-flash...');
const scenes = [
{ scene: 'work', mood: 'coffee' },
{ scene: 'work', mood: 'office' },
{ scene: 'work', mood: 'debug' },
{ scene: 'life', mood: 'gym' },
{ scene: 'life', mood: 'coffee' },
{ scene: 'life', mood: 'weekend' },
{ scene: 'mood', mood: 'happy' },
{ scene: 'mood', mood: 'tired' },
{ scene: 'mood', mood: 'focus' }
];
for (const { scene, mood } of scenes) {
try {
console.log(`Generating: ${scene}-${mood}`);
const imageUrl = await this.generateAIImage({ scene, mood });
console.log(`✓ Generated: ${imageUrl}`);
// 等待1秒,避免API限流
await new Promise(resolve => setTimeout(resolve, 1000));
} catch (error) {
console.error(`✗ Failed to generate ${scene}-${mood}:`, error);
}
}
console.log('Image library generation completed!');
}
/**
* 设置人物描述(用于自定义小跃的外观)
*/
setCharacterDescription(description: string): void {
this.characterDescription = description;
}
/**
* 获取当前人物描述
*/
getCharacterDescription(): string {
return this.characterDescription;
}
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |