npm - @thinkingai/ae-cli - Versions diffs - 1.0.1 → 1.0.3 - Mend

@thinkingai/ae-cli 1.0.1 → 1.0.3

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Files changed (170) hide show

package/skills/te-community/references/community-analyzing-theme-comment.md CHANGED Viewed

@@ -1,124 +1,124 @@
 # te-community community-analyzing-theme-comment
-> **前置条件:** 阅读 [`../te-shared/SKILL.md`](../te-shared/SKILL.md)
+> **Prerequisite:** Read [`../te-shared/SKILL.md`](../te-shared/SKILL.md)
-主题评论区深度分析：对指定帖子或视频的评论区进行深度分析，生成包含评论走势、观点分布和总结的深度分析报告。
+In-depth analysis of topic comment areas: Conduct in-depth analysis of the comment areas of specified posts or videos, and generate an in-depth analysis report including comment trends, opinion distribution, and summary.
-触发场景: 当用户需要对指定主题（帖子或视频）的评论区进行深度分析时使用此技能。
+Trigger scenario: Use this skill when users need to conduct in-depth analysis of the comment area of a specified topic (post or video).
-## 分析流程
+## Analysis process
-1. **确定目标帖子/视频**：获取 UUID、channelId 和 resourceType，或通过 `search_posts` 搜索
-2. **获取评论区数据**：调用 `get_post_detail` 获取目标内容详情和评论区
-3. **分页获取完整评论区**：循环获取所有评论，剔除灌水、纯表情等无意义内容
-4. **深度分析**：
-   - 评论走势分析（时间趋势、情感分布趋势）
-   - 观点分析（提取正面/负面观点及代表性评论）
-   - 总结分析（大盘数据、观点分析、生态分析）
+1. **Identify target post/video**: Get UUID, channelId and resourceType, or search via `search_posts`
+2. **Get comment area data**: Call `get_post_detail` to obtain the target content details and comment area
+3. **Get the complete comment area in paging**: loop through all comments and eliminate meaningless content such as irrigation and pure expressions.
+4. **In-depth analysis**:
+- Comment trend analysis (time trend, sentiment distribution trend)
+- Opinion analysis (extracting positive/negative opinions and representative comments)
+- Summary analysis (market data, opinion analysis, ecological analysis)
-## 执行流程
+## Execution process
-**步骤 1：确定目标帖子/视频**
-- 检查用户是否已明确指定需要分析的帖子或视频（UUID、channelId、resourceType）
-- 如果用户已提供以上信息，直接跳转到步骤 2
-- 如果用户未指定目标，需要使用 `search_posts` 工具搜索合适的帖子
+**Step 1: Identify Target Posts/Videos**
+- Check if the user has explicitly specified the post or video to be analyzed (UUID, channelId, resourceType)
+- If the user has provided the above information, jump directly to step 2
+- If the user does not specify a target, the `search_posts` tool needs to be used to search for suitable posts
-**步骤 2：获取评论区数据**
-- 使用 `get_post_detail` 工具获取目标内容的详细信息和评论区
-- 分页获取完整评论区（pageSize 建议 100）
-- 检查返回结果中的评论数量和 hasMore 字段
-- 如果 hasMore 为 true，继续调用获取下一页
+**Step 2: Get comment area data**
+- Use the `get_post_detail` tool to get the details and comments section of the target content
+- Get the complete comment area in pagination (pageSize recommended is 100)
+- Check the number of comments and the hasMore field in the returned results
+- If hasMore is true, continue calling to get the next page
-**评论筛选规则**：
-在获取和存储评论时，需要剔除以下无分析意义的内容：
-- **灌水评论**：重复发布相同或相似内容、刷屏内容
-- **纯表情/符号**：仅包含表情符号、符号组合（如 "？？？"、"！！！"）
-- **无实质内容**：字符数少于 3 的评论
-- **系统消息**：官方自动生成的消息
+**Comment filtering rules**:
+When obtaining and storing comments, the following content that has no analytical significance needs to be eliminated:
+- **Spoiled comments**: Repeated posting of the same or similar content, content that swipes the screen
+- **Pure emoticons/symbols**: Contains only emoticons and symbol combinations (such as "???", "!!!")
+- **No substance**: Comments with less than 3 characters
+- **System Message**: Official automatically generated message
-**步骤 3：深度分析**
+**Step 3: In-depth analysis**
-**3.1 评论走势分析**
-- 评论时间趋势：统计评论发布的时间分布，按天为单位汇总
-- 情感分布趋势：按天统计正面、中性、负面评论的比例变化
+**3.1 Comment Trend Analysis**
+- Comment time trend: statistics on the time distribution of comments published, summarized by day
+- Sentiment distribution trend: Statistics on changes in the proportion of positive, neutral, and negative comments by day
-**3.2 观点分析**
-- 对所有有效评论进行文本聚类和主题提取
-- 识别重复出现的关键词和短语
-- 合并语义相近的观点
-- 区分正面观点和负面观点
-- 每个观点包含：观点标题、观点描述、匹配评论数、代表性评论（最多 10 条）
-- **观点对立处理**：允许同一事物存在截然相反的正反观点，不要合并对立的观点
+**3.2 Viewpoint analysis**
+- Text clustering and topic extraction of all valid comments
+- Identify recurring keywords and phrases
+- Merge semantically similar views
+- Differentiate between positive and negative viewpoints
+- Each opinion includes: opinion title, opinion description, number of matching comments, representative comments (up to 10)
+- **Handling Opposing Viewpoints**: Allow diametrically opposed positive and negative views on the same thing to exist, and do not merge opposing views.
-**3.3 总结分析**
-- 大盘数据总结：评论总量、整体情感分布、评论峰值时间
-- 观点分析总结：主要正面观点、主要负面观点、观点冲突点、玩家核心诉求
-- 评论区生态分析：整体氛围、评论与内容关联度、核心用户群体特征、潜在风险点和机会点
+**3.3 Summary analysis**
+- Summary of market data: total number of comments, overall sentiment distribution, peak comment time
+- Summary of opinion analysis: main positive opinions, main negative opinions, points of conflict of opinions, core needs of players
+- Ecological analysis of the comment area: overall atmosphere, relevance of comments and content, core user group characteristics, potential risks and opportunities
-## 输出结构
+## Output structure (must be complete)
-### 一、基本信息
-- **帖子/视频标题**：标题内容
-- **来源渠道**：渠道名称
-- **发布时间**：发布日期
-- **分析时间范围**：分析的时间段
-- **有效评论数**：剔除无效评论后的数量
+### 1. Basic information
+- **Post/Video Title**: Title content
+- **Source Channel**: Channel name
+- **Published**: Release date
+- **Analysis Time Range**: The time period for analysis
+- **Number of valid comments**: The number after excluding invalid comments
-### 二、评论走势
+### 2. Comment trends
-**时间趋势**
+**Time Trend**
-| 日期 | 评论数 | 占比 |
+| Date | Number of comments | Percentage |
 |------|--------|------|
-| YYYY-MM-DD | 数量 | X% |
+| YYYY-MM-DD | Quantity | X% |
-**情感分布**
+**Emotion Distribution**
-| 日期 | 正面 | 中性 | 负面 |
+| Date | Positive | Neutral | Negative |
 |------|------|------|------|
 | YYYY-MM-DD | X% | X% | X% |
-### 三、观点分析
+### 3. Viewpoint analysis
-**正面观点**
+**Positive View**
-**观点一：[观点标题]**
-- **观点描述**：[详细描述]
-- **匹配评论数**：X 条
-- **代表性评论**：
-  - 「[评论1]」
-  - 「[评论2]」
+**Viewpoint 1: [Viewpoint Title]**
+- **Viewpoint Description**: [Detailed Description]
+- **Number of matching comments**: X
+- **Representative comments**:
+- "[Comment 1]"
+- "[Comment 2]"
-**负面观点**
+**Negative opinions (can be multiple)**
-**观点一：[观点标题]**
-- **观点描述**：[详细描述]
-- **匹配评论数**：X 条
-- **代表性评论**：
-  - 「[评论1]」
-  - 「[评论2]」
+**Viewpoint 1: [Viewpoint Title]**
+- **Viewpoint Description**: [Detailed Description]
+- **Number of matching comments**: X
+- **Representative comments**:
+- "[Comment 1]"
+- "[Comment 2]"
-### 四、总结
+### 4. Summary
-**大盘数据总结**：
-- 评论总量、整体情感分布、评论峰值时间
+**Broad market data summary**:
+- Total number of comments, overall sentiment distribution, peak comment time
-**观点分析总结**：
-- 主要正面观点
-- 主要负面观点
-- 观点冲突点
-- 玩家核心诉求
+**Summary of opinion analysis**:
+- Mainly positive points
+- Mainly negative points
+- Points of conflict of opinions
+- Players’ core needs
-**生态分析与建议**：
-- 整体氛围
-- 评论与内容关联度
-- 核心用户群体特征
-- 潜在风险点和机会点
+**Ecological Analysis and Suggestions**:
+- Overall atmosphere
+- Relevance of comments and content
+- Characteristics of core user groups
+-Potential risks and opportunities
-## 示例
+## Example
 ```bash
-# 用户输入示例
-"分析一下这个帖子的评论区"
-"帮我深度分析这个视频的用户反馈"
+# User input example
+"Analyze the comment area of this post"
+"Help me deeply analyze the user feedback of this video"
 ```

package/skills/te-community/references/community-character-analysis.md CHANGED Viewed

@@ -1,71 +1,71 @@
 # te-community community-character-analysis
-> **前置条件:** 阅读 [`../te-shared/SKILL.md`](../te-shared/SKILL.md)
+> **Prerequisite:** Read [`../te-shared/SKILL.md`](../te-shared/SKILL.md)
-角色专项深度分析：深度剖析特定游戏角色的社区生态、玩家情感动因、核心诉求，输出结构化的《角色专项深度分析报告》。
+Role-specific in-depth analysis: In-depth diagnosis for a single role, instead of regular daily/weekly reports, the "Role-specific in-depth analysis report" is output.
-触发场景: 当用户要求"深度分析某角色"、"角色舆情洞察"、"拆解某角色讨论情况"时触发。
+Trigger: Use this when the user asks for "in-depth analysis of a certain character", "character sentiment insight", "deep-dive analysis of of a certain character's discussion", and "character analysis".
-## 分析流程
+## Analysis process
-1. **`get_sentiment_overview`**：获取该角色在正负面讨论下的核心词云，定调当前总体口碑
-2. **`get_hot_topics`**：确认该角色是否在社区形成了聚集性的热门话题
-3. **`search_posts`**：以"角色名 + 核心争议词"搜索近期高热帖子
-4. **`get_post_detail` & `get_comments_summary`**：提取帖子核心观点和评论区情感反馈
+1. **`get_sentiment_overview`**: Get the core word cloud of the character under positive and negative discussions to set the tone of the current overall reputation
+2. **`get_hot_topics`**: Confirm whether the character has formed a hot topic in the community
+3. **`search_posts`**: Search for recent hot posts using "character name + core controversial words"
+4. **`get_post_detail` & `get_comments_summary`**: Extract the core views of the post and emotional feedback in the comment area
-## 实体提取与推演规则
+## Entity extraction and deduction rules
-- 必须从用户输入中提取【核心分析角色】（如"角色"、"角色A"）
-- 如果用户没说，请明确追问"请问您希望深度分析哪位角色？"
-- 若用户未指定时间，**自动以当前系统时间往前推 7-14 天作为查询周期**
+- [Core Analysis Role] must be extracted from user input (such as "Role", "Role A")
+- If the user doesn't say anything, please clearly ask "Which character do you want to analyze in depth?"
+- If the user does not specify a time, **automatically push forward 7-14 days from the current system time as the query period**
-## 核心约束
+## Core Constraints
-- **拒绝假大空**：必须写出"为什么觉得低（对比了谁？什么输出环境？机制哪里卡手？）"，并附上具体帖子或评论原声佐证
-- **严格区分预警与风控**：
-  - 预警（Alerts）：游戏内体验问题（机制Bug、数值设计不合理、剧情吃书崩坏）
-  - 风控（Risk Control）：触碰现实底线的社区安全问题（涉及该角色的引战谩骂、黑产交易、内鬼解包泄露）
+- **Reject fake claims**: Must explain "why" and give evidence of posts/comments.
+- **Strictly distinguish between early warning and risk control**:
+- Alerts: experience risks (mechanics, numerical values, plot, art, performance)
+-Risk Control: risk of violation (causing war, illegal production, leakage, rumors)
-## 输出结构
+## Output structure
-### 一、角色当前生态与社区印象
-用一段极具洞察力的文字，概括该角色目前在玩家心智中的真实定位
+### 1. Character’s current ecology and community impression
+Use a very insightful paragraph to summarize the character’s current true position in the player’s mind.
-### 二、讨论热度与情感画像矩阵
+### 2. Discussion popularity and emotional portrait matrix
-| 维度 | 数据/描述 |
+| Dimensions | Data/Description |
 |------|----------|
-| 情感基本盘 | 正面/中性/负面占比 |
-| 舆论主阵地 | 主要讨论渠道 |
-| 高频词共现 | 核心关键词 |
+| Emotional fundamentals | Positive/neutral/negative ratio |
+| Main forum for sentiment | Main discussion channel |
+| High frequency word co-occurrence | Core keywords |
-### 三、玩家核心诉求与焦点深度拆解
+### 3. In-depth deep-dive analysis of of players’ core needs and focus
-**✨ 核心价值与高光体验**
-- **玩家心声提炼**：[总结]
-- **典型原声/支撑案例**：「[真实评论]」
+**✨ Core Values and Highlight Experience**
+- **Extraction of player voices**: [Summary]
+- **Typical original voice/support case**: "[Real comments]"
-**🧨 核心痛点剖析**
-- **深层诉求拆解**：[分析]
-- **典型原声/支撑案例**：「[真实评论]」
+**🧨 Analysis of core pain points**
+- **Dismantling of deep demands**: [Analysis]
+- **Typical original voice/support case**: "[Real comments]"
-### 四、运营策略
-- **舆论风向引导**：具体建议
-- **版本/优化前瞻规划**：改进方向
+### 4. Operation strategy
+- **Public Opinion Direction Guidance**: Specific Suggestions
+- **Version/Optimization Forward Planning**: Direction for improvement
-### 五、预警内容
+### 5. Warning content
-**⚠️ 预警内容**
+**⚠️ Warning content**
-| 危险等级 | 预警触发项 | 社区蔓延度 | 核心表现 | 建议跟进动作 |
+| Danger level | Warning triggers | Community spread | Core performance | Suggested follow-up actions |
 |---------|-----------|-----------|---------|-------------|
-| 高/中/低 | 具体问题 | 影响范围 | 具体表现 | 处理建议 |
+| High/medium/low | Specific problems | Scope of impact | Specific performance | Handling suggestions |
-## 示例
+## Example
 ```bash
-# 用户输入示例
-"深度分析一下角色这个角色"
-"帮我拆解角色A的讨论情况"
-"分析下XX角色的舆情"
+# User input example
+"A deep analysis of the role"
+"Help me break down the discussion of character A"
+"Analyze the sentiment of XX character"
 ```

package/skills/te-community/references/community-daily-report.md CHANGED Viewed

@@ -1,82 +1,85 @@
 # te-community community-daily-report
-> **前置条件:** 阅读 [`../te-shared/SKILL.md`](../te-shared/SKILL.md)
+> **Prerequisite:** Read [`../te-shared/SKILL.md`](../te-shared/SKILL.md)
-每日快报：生成单日大盘、突发舆情与安全合规风控的短平快《用户运营每日快报》，支持环比对比（T-1 与 T-2）。
+Daily brief: Generate a concise "User Operations Daily Brief" covering daily macro trends, breaking sentiment signals, and safety/compliance risks, with T-1 vs T-2 comparison.
-触发场景: 当用户要求"生成日报"、"昨日战报"或"每日舆情总结"时触发。
+Trigger: Use this when the user asks for "generate daily report", "yesterday's battle report" or "daily sentiment summary".
-## 分析流程
+## Roles and Goals
-1. **确定时间参数**：目标日(T-1)和对比日(T-2)
-2. **调用 MCP 工具**：
-   - `get_overview_metrics`（两次）：获取 T-1 和 T-2 的发帖/反馈总数，计算环比涨跌幅
-   - `get_sentiment_overview`：获取目标日情感分布
-   - `get_daily_summary`：获取目标日核心事件摘要
-   - `search_posts`：搜索目标日高热度话题原帖
-   - `get_post_detail` + `get_comments_summary`：提取真实用户原话
-   - `get_risk_content`：获取安全合规风控数据
+You are the "User Operations Director + Public Opinion Monitoring Expert". The core task is to quickly answer three things:
+1) Whether the overall trend on the target date rose or fell relative to the previous day; 2) What is the most important focus of sentiment on the day; 3) Are there any risks that need to be dealt with immediately?
-## 时间推演与动态话术规则
+## Time deduction and speech rules
-1. **用户明确指定日期**（如："出一下3月17日的日报"）：
-   - 目标日为指定日期，对比日自动推算为前一天
-   - **话术红线**：报告正文中**绝对禁止出现"今日"、"昨日"、"前天"等相对时间代词**，必须使用具体日期或客观代词（如："当日声量激增"、"对比上日"）
-2. **用户未明确指定日期**（如只说"生成日报"）：
-   - 立即获取当前系统时间，**自动以"昨天"作为目标日(T-1)**，前天为对比日(T-2)
-   - 此时允许在报告中自然使用"昨日大盘"、"今日重点"等常规表述
+1. **The user clearly specifies the date** (for example: "Publish the daily report on March 17"):
+- The target date is the specified date, and the comparison date is automatically calculated as the previous day.
+- **Red line of speech**: In the text of the report, relative time pronouns such as "today", "yesterday" and "the day before yesterday" are absolutely prohibited**, and specific dates or objective pronouns must be used (such as: "Surge in volume on that day", "Compared to the previous day")
+2. **The user did not explicitly specify the date** (such as only saying "generate daily report"):
+- Automatically use yesterday as the target day (T-1) and the day before yesterday as the comparison day (T-2)
+- Allows the use of relative terms such as "yesterday/today's highlights"
-## 核心约束
+## Required tool chain
-- **极简主义，拒绝长篇**：重点描述"增量信息"和"突发异常"
-- **预警与风控的绝对隔离**：产品Bug、UI吐槽、活动反馈归为"业务体验"；黑产、涉政、黄赌毒、违规广告写进"合规风控"
-- **用户原声绝对真实**：必须 100% 来源于真实文本，**绝对禁止捏造或臆想用户发言**
+1. `get_overview_metrics` (called twice: T-1 and T-2)
+2. `get_sentiment_overview` (target day sentiment distribution)
+3. `get_daily_summary` (target day event summary)
+4. `search_posts` (locating core hot posts)
+5. `get_post_detail` + `get_comments_summary` (extract the real original sound)
+6. `get_risk_content` (compliance risk)
-## 输出结构
+## Core Constraints
-### 一、核心概览
-- **摘要总览**：一句话概括当日核心事件
-- **数据概览**：关键指标环比变化
-- **当日热词**：高频关键词
+- **Minimalism, rejecting long forms**: focus on describing "incremental information" and "sudden anomalies"
+- **Absolute isolation of early warning and risk control**: Product bugs, UI complaints, and event feedback are classified as "Business Experience"; illegal products, politics, pornography, gambling and drugs, and illegal advertising are written into "Compliance Risk Control"
+- **The original voice of the user is absolutely authentic**: it must be 100% derived from real text, **It is absolutely prohibited to fabricate or conjecture what the user said**
-### 二、单日核心看板
+## Output structure
-| 指标类别 | 数据 | 环比变化 |
+### 1. Core Overview
+- **Summary Overview**: Summarize the core events of the day in one sentence
+- **Data Overview**: Month-on-month changes in key indicators
+- **Hot words of the day**: high-frequency keywords
+### 2. Single-day core dashboard
+| Indicator categories | Data | Month-on-month changes |
 |---------|------|---------|
-| 大盘总声量 | 具体数值 | ↑/↓ X% |
-| 当日情感大盘 | 正面/中性/负面占比 | 变化趋势 |
-| 主力声量阵地 | 各渠道分布 | 渠道对比 |
+| Total sound volume of the market | Specific value | ↑/↓ X% |
+| Sentiment of the day | Positive/neutral/negative proportion | Changing trends |
+| Main volume position | Distribution of various channels | Channel comparison |
-### 三、核心焦点与用户洞察
+### 3. Core Focus and User Insights
-**核心焦点**：
-- 焦点一：[事件描述]
-  - **用户反馈**：[反馈总结]
-  - **典型原声**：「[真实用户评论]」
+**Core Focus**:
+- Focus 1: [Event Description]
+- **User Feedback**: [Feedback Summary]
+- **Typical original voice**: "[Real user comments]"
-**用户情绪总结**
+**User Sentiment Summary**
-| 情绪倾向 | 主要触发点 | 可能的影响 |
+| Emotional Tendencies | Main Trigger Points | Possible Impact |
 |---------|-----------|-----------|
-| 正面/负面/中性 | 具体原因 | 影响分析 |
+| Positive/Negative/Neutral | Specific reasons | Impact analysis |
-### 四、风控提示与运营建议
+### 4. Risk control tips and operational suggestions
-**安全合规风控**
+**Safety Compliance Risk Control**
-| 风控类型 | 发现数量 | 主要表现 | 处置建议 |
+| Risk control type | Number of findings | Main performance | Disposal suggestions |
 |---------|---------|---------|---------|
-| 具体类型 | 数量 | 具体表现 | 处理建议 |
+| Specific type | Quantity | Specific performance | Treatment suggestions |
-### 五、运营建议与未来节点建议
-- 短期建议
-- 中长期关注点
+### 5. Operation suggestions and future node suggestions
+- short term advice
+- Mid- to long-term focus
-## 示例
+## Example
 ```bash
-# 用户输入示例
-"生成日报"
-"出一下3月17日的日报"
-"昨日战报"
+# User input example
+"Generate daily report"
+"Publish the daily report for March 17th"
+"Yesterday's Battle Report"
 ```