CYBERDELIA FOIA INTAKE INDEXER v0.1.0
=====================================

Purpose
-------
Build a reproducible first-pass inventory of a directory containing FOIA,
declassified, or other public-record source files without modifying the source.

The tool emits:
  * manifest.json   - complete inventory + integrity fields
  * inventory.csv  - flat spreadsheet-friendly inventory
  * duplicates.json - exact duplicate groups by SHA-256

Requirements
------------
Python 3.10+; standard library only.

Usage
-----
python foia_intake_indexer.py /path/to/source --out ./intake-output

Optional exclusions can be repeated:
python foia_intake_indexer.py ./records --ignore thumbnails --ignore temp

Evidence boundaries
-------------------
* SHA-256 proves byte identity, not truth.
* File modification time is filesystem evidence, not automatically creation time.
* MIME type is guessed from the filename extension and is not forensic file typing.
* Exact duplicates are byte-for-byte duplicates only. Near-duplicate/OCR comparison is
  deliberately outside v0.1.0.
* The utility does not alter, rename, delete, or move source records.

Intended workflow
-----------------
1. Preserve original source files.
2. Run the intake indexer.
3. Store the generated manifest beside acquisition provenance.
4. Use duplicate groups to reduce redundant review without deleting originals.
5. Build OCR, annotations, timelines, and AI-assisted derivatives separately.

License
-------
See LICENSE.txt.
