ℹ
Purpose & Capability
The name/description match the included code: a Python/pandas-based data cleaner. The README and SKILL.md mention batch processing and a batch_clean.py example, but no batch_clean.py is included in the package — a minor inconsistency. Overall, required tools (pandas/openpyxl) are appropriate for the stated purpose.
ℹ
Instruction Scope
Runtime instructions and examples only run local Python scripts on local files and request installing pandas/openpyxl. They do not instruct reading unrelated system files or exporting data to external endpoints. One mismatch: the SKILL.md and README call this an 'AI-driven' tool, but the provided script is heuristic/deterministic (no external model/API calls).
✓
Install Mechanism
No install spec; dependencies are installed via standard pip (pandas, openpyxl) as documented. No downloads from untrusted URLs or archive extraction are present.
✓
Credentials
The skill requests no environment variables, no credentials, and no special config paths — proportional to a local data-cleaning tool.
✓
Persistence & Privilege
Skill is not marked always:true and is user-invocable (normal). It does not request persistent system-wide privileges or modify other skills/config.
Assessment
This package appears to be a straightforward local data-cleaner and does not request credentials or make network calls, so risk is low. Before installing/running: 1) note that README/SKILL.md mention batch_clean.py but that file is missing — batch-processing example is incomplete; 2) the script is deterministic (pandas/regex rules), not actually calling any external AI service despite the 'AI-driven' wording; 3) there are small code/documentation mismatches (e.g., --fix-date handling in the CLI and the function signature) but these are bugs, not malicious behavior; 4) sanitize or review outputs if your data may contain untrusted spreadsheet formulas: the tool writes CSV/Excel but does not sanitize leading characters (e.g., =,+,-,@) which can be a CSV/Excel formula-injection risk when opening in spreadsheet apps; 5) run it first on non-sensitive test data or in a sandbox, and inspect the generated _cleaned files to confirm behavior. If you need batch processing, request the missing batch script from the author or implement your own wrapper.
