Casework and Evidence

Casework and Evidence

Case Flow

1. Create a case or use the local inbox.
2. Add a manual observation or validated JSON import with its source URL and capture time.
3. Attach source records, notes, tags, screenshots, or explicit local image evidence.
4. Review profile changes, relationship candidates, timeline events, and graph connections.
5. Confirm, reject, or qualify review items.
6. Export the case as JSON, PDF, or an opt-in encrypted EvidenceVault package.

Traceability

Observations are fingerprinted locally. Relationship candidates retain their evidence observation. Profile changes and identity hypotheses retain their basis, state, confidence, and evidence reference. Timeline and graph views link back to the source record wherever possible.

Evidence Types

• Captured observations and external source links.
• Local image evidence and screenshot references.
• Derived OCR text and relative-depth artifacts.
• Notes, tags, profile snapshots, and review decisions.

Derived data does not replace the original local evidence. OCR and depth runs retain the model profile, timestamp, and result metadata. The evidence vault can package source files and completed depth derivatives with hashes, encryption, and a signature.

Commands

The local CommandDeck supports casework commands such as:

/find posts --query "term" --limit 20
/profile connections @account
/graph path @account_a @account_b --max-hops 4
/timeline build
/export case --format json

Commands are allowlisted and operate on the local case database only.