Remark42

Unlocking Insights: TikTok's New Mini Game Analytics Suite

TikTok's Developer Portal documents a Mini Game Analytics suite that centralizes daily performance metrics for mini games across monetization, promotion and operations. The guide lists analytics modules accessible on an app's Developer Portal page — including Overview, In‑App Purchases (IAPs), In‑App Ads (IAAs), For You feed ads, For You feed cards, Users and Performance — and describes available metrics, filters (date range, region, OS, ad placement) and download/graphing capabilities. Monetization analytics are deduplicated and include user- and order-level IAP views (GMV, orders, ARPPU) and ad metrics (requests, impressions, clicks, click-through rate, eCPM, revenue). For You feed ad reporting is by calendar day and TikTok recommends matching TTAM to calendar-day reporting for consistency. Operational analytics cover behavior, retention, profile, device and performance (launch speed, success rate).

China's Tech Giants Race to Dominate OpenClaw AI

China’s tech firms and local governments have rapidly adopted OpenClaw, an open-source AI agent, with major companies launching localized or integrated versions and offering installation support to consumers. Tencent released an OpenClaw-based product suite compatible with WeChat; Zhipu AI published a local OpenClaw build preinstalled with 50+ skills; ByteDance’s Volcano Engine launched a browser-accessible 'ArkClaw' variant. Usage of OpenClaw in China has surpassed the U.S. per SecurityScorecard, and startups, retailers and districts are offering paid deployment services and subsidies (some up to 10 million yuan) to spur development. The trend is lowering costs to run agentic AI, boosting interest in domestic LLMs, accelerating monetization efforts, and raising privacy and security concerns due to agents’ broader access to user data and systems.

Revolutionary Open-Source Server Transforms Document Translation for AI

Bluente announced an open-source MCP (Model Context Protocol) server that embeds format-preserving document translation into MCP-compatible AI environments such as Claude Desktop and Cursor. The server lets AI agents translate documents across 120+ languages while keeping original formatting (tables, numbering, tracked changes) and supports common file types including PDF, DOCX, XLSX, PPTX and images with integrated OCR for scanned docs. The MCP server is available on GitHub under the MIT license, runs on Node.js (v20+), communicates with AI clients over stdio and Bluente APIs over HTTPS, and enforces enterprise security measures (zero data retention, end-to-end encryption, automatic deletion). Bluente positions the release as a way to reduce context-switching and simplify developer pipelines for translation workflows.