Sensitive data release

[sebbacon]

Tooling to support safe but relatively painless release of potentially identifiable data according to our published policy.

This may include:

• Command-line tools to double-check for patterns that suggest sensitive data might be accidentally released
• Automated algorithms for measuring potential disclosivity (e.g. k-anonymity)
• Graphical tools for comparing changes in released outputs over time
• Cryptographic signing of known-safe filtering/redaction code
• Separate local-only git repos for auditing and managing release workflow
  • Graphical workflow tools for this (e.g. "Anna reviewed this on date X and sent back to Bob for changes", etc)
• Automatic publication of known-safe data