- Home
- Skills
- Writing & Content
- mama-message — Family Chat Reports & Reply Drafts
More screenshots
Works with the AI Agents you already use
mama-message — Family Chat Reports & Reply Drafts
Turn family chat exports into a report, two shareable cards, and a message you can send. Chinese-first, with local Python statistics and AI-written reply drafts.
Free
mama-message — Family Chat Reports & Reply Drafts
Example session with this skill installed
请分析我上传的 sample-messages.json,生成我和妈妈的聊天报告、两张卡片,并给我一条今晚可以发给她的话。
Demo input: the project's synthetic sample conversation. All messages are fictional; no real family chat is published here. For this reproducible example, use 2026-08-30 as the report date.
- Read your context and instructions
- Compiled the mama-message
- Generated the document
Demo: 你和妈妈的半年
This example uses fictional messages generated by the project. The figures below come from its local Python scripts; the Chinese reply text comes from the included sample narrative.
聊天概况
| 项目 | 结果 |
|---|---|
| 记录范围 | 2026-03-01 至 2026-08-28 |
| 消息总数 | 294 条 |
| 妈妈发来的消息 | 192 条 |
| 我发出的消息 | 102 条 |
| 有聊天的天数 | 119 天 |
| 妈妈先开口 | 86 次,占 67.2% |
| 我先开口 | 42 次 |
| 我回复她的中位间隔 | 1 小时 30 分 |
三个指数的说明
- 牵挂:她发起了 67% 的对话,有 28 条是早上 8 点前发的、51 条是晚上 10 点以后。
- 短回复:你 93 条回复里有 52 条不超过三个字。
- 没回之后:你没回的 59 次里,有 51 次是她自己先开口了下一次。
这些数字描述这份模拟记录里的交流模式,不用于判断现实中的家庭关系。
可以直接发给她的话
妈,我用了个软件,它把咱俩这半年的聊天记录算了一遍:你发了我 192 条,我回了 102 条,还有 59 条我压根没回。它说我欠你一个电话。今晚 8 点打给你,你别做饭了,等我电话。
帮你回
记录中的消息:08-14 “最近腰不太好 医生说要少干活”
回复草稿:“妈,腰现在怎么样了?医生让你少干活你就少干,家里的活等我回去干。”
在真实使用中,草稿会根据用户的记录和补充信息调整;没有记录支持的事实不应加入。
输出文件
The rendering step creates
mama-message-report.html: the full report with statistics and Chinese narrative.mama-message-card-bill.html: a shareable conversation summary card.mama-message-card-send.html: the message card for the family member.
Open the HTML files in a browser to view them or capture them as images. The skill drafts messages; the user decides what to send.
The bundled scripts use the Python standard library and do not send chat data over the network. Chat records supplied to an online agent are handled according to that agent provider's policies.
mama-message-report.html
PDF · document
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
mama-message helps you look through conversations with your mom, dad, grandparents, or other family members and find a useful next reply.
Start with a WeFlow JSON export, a supported CSV/TXT export, pasted text, or screenshots. The skill identifies the two participants, calculates message counts and reply patterns, and helps your agent write a grounded report using the actual conversation. It also creates two HTML cards and a ready-to-send message with a concrete time to call or catch up.
Features include:
- Message counts, conversation initiators, reply times, short replies, and unanswered messages.
- Chinese-language report text and reply drafts grounded in your supplied messages and computed statistics.
- A full HTML report and two shareable cards.
- Optional reply practice and comparisons across family or friend conversations.
- Local redaction tools; transfers are counted without recording amounts.
Requirements: an AI agent able to read files and run Python 3.8 or later. The included scripts use the Python standard library and make no network requests. Screenshot input needs an agent that can read images. Exporting WeChat history is a separate user-controlled step; this skill does not decrypt WeChat databases.
The skill is designed primarily for Chinese conversations and Chinese output. HTML cards can be opened in a browser and captured as images. Records provided to an online AI agent are processed according to that agent provider's data policies. The reports describe communication patterns and are not psychological assessments.
Free and open source under the MIT license.
Chinese-first instructions, writing rules, and report output. This skill does not export or decrypt WeChat databases; users supply their own records. WeFlow exports may lack reliable voice-message durations. Messages without timestamps are counted but excluded from time-based metrics. Very small or incomplete datasets limit interpretation. HTML cards need a browser to view or capture as images. Scripts are local, but records supplied to an online AI agent follow that provider's data policies.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
1 use
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 1 month ago
- Free with an Agensi account
Requires
Permissions
Reads only chat records supplied by the user. Runs bundled Python scripts to validate records, calculate statistics, redact sensitive fields, and write JSON/HTML reports in the user's chosen working directory. No connected service account, environment variable, or network access is required by the scripts. Opening the HTML output in a browser is optional. Records given to an online AI agent are subject to that provider's data policies.