Multi Source Data Cleaner

Security checks across malware telemetry and agentic risk

Overview

The code and instructions mostly match a data-cleaning tool, but documentation makes conflicting claims about network usage (saying processing is local) while the skill clearly includes optional AI features and Feishu export that will send data externally — this mismatch and the external-API/credential surface warrant caution.

This package appears to be a legitimate multi-source data cleaner, but pay attention to two important points before installing or running it with real/personal data: - Network/data exfiltration: AI features (MiniMax / DeepSeek) and Feishu export will send data to external services. Despite a README line saying "all processing is local," the code and SKILL.md show optional calls that require API keys or Feishu tokens. Do not enable AI classification or Feishu export if you must keep data strictly on-premise. - Secrets and tokens: DATA_CLEANER_API_KEY and Feishu folder/open_id tokens are sensitive. Only provide them if you trust the endpoints and have reviewed the code paths that use them (look at classifier.classify_with_ai, output.DataExporter.to_bitable and to_feishu_doc). - Audit the network calls: If you need stronger assurance, inspect the omitted parts of the code (classify_with_ai and DataExporter implementations) to see exact API endpoints, domains, and any logging/exfiltration behavior. Run the tool in a sandbox or restricted network environment first. - Runtime environment: The skill is Python-based and lists pip dependencies; ensure those are installed in an isolated venv. The skill persists usage to /tmp/data_cleaner_state.json — clear or relocate that if needed. If you want, I can (1) scan the remaining truncated files for explicit HTTP calls and endpoints (classify_with_ai, output.py, reporter.py), or (2) produce a short checklist of code locations to review before providing any API keys or tokens.

SkillSpector

By NVIDIA

SkillSpector findings are pending for this release.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal