
, also called Singwork on its site...
Instead of scanning a sheet for anomalies every day, the assistant reads first and says what to handle. The daily brief names the issue, the reason, and the first action; the page examples cover stock, add-to-cart, and ads. Chat entry can connect Lark, Telegram, or WhatsApp. Runtime scheduling puts Claude Code and OpenCode in one arrangement. Conversation models can be switched per task. The comparison table says it remembers business context and gives an executable step, rather than starting every chat from scratch. It fits store owners and operators who want a daily data check, a file or store connection.
SingClaw, also called Singwork on its site, is a desktop data agent with memory for Mac and Windows. After you drop in a file or connect store data, it reads the data and returns a daily business brief: what to handle first, why, and the action to take. Examples on the page include stock risk, an add-to-cart change, and an advertising anomaly. It can switch among conversation models, connect Lark, Telegram, and WhatsApp, and schedule runtimes such as Claude Code and OpenCode. Insights combine structured parsing, visualization, and recommendations, and the page says the data stays under your control.
Difficulty: Beginner Friendly
Daily business brief
After a file or store connection, it reads the data and lists what to handle today, why, and the one action to take first.
Switch models per task
Several mainstream conversation models are built in and can be switched per task for different analysis or writing needs.
Chat app connections
Lark, Telegram, and WhatsApp can be connected so group or message workflows reach this desktop agent.
Several coding runtimes
The page says it can schedule Claude Code, OpenCode, and other runtimes together in one arrangement.
Daily store anomalies
After connecting store data, an owner reads the brief for stock, conversion, or ad anomalies and decides what to handle immediately.
Drop a file, get a conclusion
An analyst drops a spreadsheet into the desktop app and reads the flagged issues and suggested actions without building a dashboard first.
Briefs inside chat apps
An operator connects Lark, Telegram, or WhatsApp and receives the data conclusions inside the chat tool they already use.
The page reduces the start to three steps. Drop in a file or connect store data; it says this takes about three minutes with no extra setup. The assistant then reads the data, so you do not build a dashboard first. The daily brief says what is wrong, why, and the one action to take first. Examples include stock risk, an add-to-cart change, and an ad anomaly. Mac and Windows downloads are both listed, and the page says you can start free.
The product title and comparison table both stress memory. In that table, a general assistant is a single session, while this one keeps building business context and uses your industry terms. The page says the brief shows up before you open the laptop and includes an executable step. The privacy row contrasts staying under your control with a general assistant that uploads data to the cloud. You can still switch conversation models per task; memory and model choice are separate.
The integration section lists Lark, Telegram, and WhatsApp. Code runtimes listed are Claude Code and OpenCode, scheduled together. Data results combine structured parsing, visualization, and insight rather than only raw numbers. Downloads are split into Mac and Windows. The page says the download is free and you can start free, but it does not list paid quotas, so any later limit has to be checked on the live site.
Real reviews and feedback from users