Perfect for data analysts, journalists, anyone dealing with CSV exports from multiple systems. Try it, star it, or contribute:
https://github.com/Intellirim/csv-surgeon (7/7)

Install in one line: `pip install csv-surgeon`

Works in data pipelines, handles files >1GB with stream processing, optional PII sanitization mode. (6/7)

Want to see what's wrong first? `csv-surgeon analyze file.csv` gives you diagnostics: encoding confidence, delimiter analysis, structural issues with line numbers. (5/7)

Before: broken.csv with 127 malformed rows, encoding issues, quote problems.
After: `csv-surgeon repair broken.csv` → clean, valid CSV in seconds. (4/7)

It handles the nasty stuff: UTF-8/Latin-1/CP1252 encoding detection with confidence scoring, unclosed quotes with context-aware repair, embedded linebreaks that split records across lines. (3/7)

csv-surgeon uses statistical analysis to detect your delimiter automatically. No more guessing if it's comma, semicolon, or tab. Coefficient of variation finds the most consistent one. (2/7)

Try it out, star the repo, or contribute! Built for developers who need GDPR/HIPAA compliance without expensive enterprise DLP tools. https://github.com/alphaoftech/pii-guard (8/8)

Integrate into pre-commit hooks to block commits with PII. Add to GitHub Actions for automated security checks. Use in your API middleware to sanitize data before LLM calls. 2 lines of code. (7/8)

Install in one line:
pip install pii-guard

Scan a file:
pii-guard scan input.txt

Mask and save:
pii-guard scan --mask partial --output clean.txt input.txt

Works with stdin for pipelines too. (6/8)

Multiple masking options: full redaction, partial masking (***-**-1234), hash replacement, or tokens. Output as JSON for automation or human-readable reports. Configurable confidence thresholds. (5/8)

Best part? It runs entirely locally at 10MB/sec. Zero external API calls means your data stays private. Perfect for CI/CD pipelines or pre-processing LLM inputs. No telemetry, no cloud dependencies. (4/8)

It detects 50+ PII types: SSNs, credit cards, emails, phone numbers, passports, medical IDs, and API keys from AWS, OpenAI, Stripe, GitHub, and more. Each pattern has custom validation logic like Luhn checksums for cards. (3/8)

pii-guard is a context-aware PII detector for LLM pipelines. Unlike regex-only tools with 40%+ false positives, it analyzes surrounding text to distinguish '555-1234' in version strings from actual phone numbers. 60% fewer false alarms. (2/8)