
from raccoon-dataanalysis-skill177
Remote data-analysis and visualization skill that uploads user files to the Raccoon API to produce charts, code-interpreter results, and multi-turn analysis ses
Raccoon Data Analysis connects an agent to a remote code-interpreter and visualization service. Rather than processing files locally, this skill uploads user-provided data (Excel, CSV, etc.) to the configured RACCOON_API_HOST and uses the remote API to run analysis, generate visualizations (charts, images), and return artifacts for download. It provides a structured, privacy-aware workflow that requires explicit user consent before any upload.
Use this skill when a user asks you to analyze or visualize a dataset and is okay with uploading files to an external service. Typical triggers: 'analyze this Excel', 'plot a chart from my CSV', 'generate insights from the dataset'. Do not trigger if the user forbids external uploads or asks for local-only processing.
Agents that can run Python scripts and call external HTTP APIs (e.g., Copilot/Codex/Claude Code or any agent with shell/python execution ability).
Raccoon Data Analysis is a skill that uploads user files (Excel, CSV) to SenseTime's xiaohuanxiong.com API for remote AI-driven analysis and visualization. The single script (main.py) ran successfully, providing a well-structured CLI with analyze, auth-check, and example subcommands. The code is cleanly organized with proper error handling, retry logic, and SSE streaming support.
Well-structured skill with clear privacy documentation and user consent requirements in the SKILL.md. Explicitly prohibits local data analysis, routing everything through the remote API. The skill is specific to the Chinese market and SenseTime's ecosystem. No references/ directory was bundled despite being mentioned in SKILL.md. No type hints on functions.