Agent Skills · Data Analysis
Curated skills for Claude, Cursor and other AI agents. Copy one prompt to install.
Data-analysis skills claim to turn a table, a log, or an export into a summary, a chart, or a prediction. The legitimate use is a file you are allowed to read, processed in a way you can recompute. The common misuse is a personal chat export or a customer dump sent to whichever model the agent prefers.
Ask where the file goes. A card that says "fully local" is making a claim you can test by reading the prompt. If the prompt tells the agent to call a hosted model or upload the file, it is not local, whatever the adjective was. When not to install a skill treats other people's messages as a stop condition. That includes group chats you exported because you were curious.
A browser-launcher skill on this shelf is automation that happens to feed an analysis. It is closer to a web plugin than to a statistics library. Read which browser profile it opens. Your daily profile contains sessions you do not want a prompt to drive.
The output should include the numbers it computed and the rows they came from. A paragraph of "insights" with no table is not an analysis you can defend. If you need a model that can hold a large table in context, that is a model listing decision, not a reason to stack three analysis skills.
Do not install a predictor for people, hiring, credit, or medical risk from this category. The directory will contain experiments. An experiment is not a process you point at a person.
Keep the source file next to the summary. If you cannot recompute one figure from that file in a minute, the summary is not finished. Delete the insight paragraph and ask for the table again before you paste the result into a decision.
A fully local-run WeChat chat analysis tool with MBTI personality inference, sentiment analysis, risk detection, relationship health assessment, and conversation prediction, emphasizing privacy.
A universal base component that automatically launches the Edge browser and connects the BrowserSkill extension, ensuring readiness for web scraping and browser automation tasks.
Turn pasted or queried Xiaohongshu notes into a research memo covering title patterns, structure, quoted comments, and follow-up angles; public lookups are optional and confirmed before paid execution.
An A-share market monitor using Sina Finance API for real-time quotes, price and percentage alerts, daily logging, and scheduled checks.
Scrapes JD search results for product titles, prices, sales, and stores, then exports verified data to UTF-8 CSV.
Performs local, paragraph-level cross-document similarity checks and produces an HTML report with revision suggestions.
Reads Excel/CSV files, cleans and analyzes them, then generates a charted Excel summary and interactive HTML dashboard.
Analyzes product VOC and brand reputation events across public news, community, and social pages, delivering Markdown, Excel, and optional HTML reports.
Batch-classify pasted text or CSV rows with preset and custom templates, outputting structured labels and confidence scores for data organization and annotation workflows.
Query Douyin hot searches and works with SocialDataX API or MCP tools, supporting keyword, sorting, time range, duration, content type, and pagination.
Fetch the GitHub Trending leaderboard and generate fact-grounded project analysis, trend hypotheses, and directional insights.
Query Douyin creator videos, image/text posts, and short-drama series by sec_user_id or profile URL with paginated list retrieval.