
Claim Guardians
ClaimsCalQL8ter
AI Status Story Review
1: 33-88J6-21T · Mary Weaver
Review the customer-facing story before it is published. This screen keeps the raw Claude run history, story versions, and usage logs together so staff can check the narrative against the claim facts.
AI Review
This is the staff review screen. Generate a fresh draft, inspect the latest AI run, then approve or publish only after the story reads cleanly for the customer.
Story Versions
3
drafts, approvals, and published versions on file
AI Runs
2
status summary executions logged for this claim
Latest Model
claude-sonnet-4-20250514
sonnet tier · completed
Story Versions
Newest first. Only approved drafts should be published.
Version 3
draftThis claim needs internal review before the next client update is released.
Next step: Record the primary customer contact.
Generated: Jun 15, 2026, 1:39 AM
Model: claude-sonnet-4-20250514
Prompt: client-summary-v1
Published: —
Review flags
Version 2
draftThis claim needs internal review before the next client update is released.
Next step: Record the primary customer contact.
Generated: Jun 15, 2026, 1:38 AM
Model: claude-sonnet-4-20250514
Prompt: client-summary-v1
Published: —
Review flags
Version 1
publishedWe are waiting on a client item before the claim can move forward.
Next step: Record the primary customer contact.
Generated: Jun 12, 2026, 11:01 PM
Model: deterministic-fallback
Prompt: client-summary-v1
Published: Jun 12, 2026, 11:03 PM
Latest AI Runs
status summary
completedTrigger: manual
Model: claude-sonnet-4-20250514
Started: Jun 15, 2026, 1:39 AM
Completed: Jun 15, 2026, 1:39 AM
status summary
completedTrigger: manual
Model: claude-sonnet-4-20250514
Started: Jun 15, 2026, 1:38 AM
Completed: Jun 15, 2026, 1:38 AM
Claude Usage
claude-sonnet-4-20250514
error
Input tokens: 0
Output tokens: 0
Estimated cost: $0.0000
Logged: Jun 15, 2026, 1:39 AM
claude-sonnet-4-20250514
error
Input tokens: 0
Output tokens: 0
Estimated cost: $0.0000
Logged: Jun 15, 2026, 1:38 AM
Document Intake Review
This is where staff verifies AI-extracted claim fields before those suggestions become trusted intake signals. Approve clean candidates, reject bad ones, and keep the claim ledger protected from duplicate or low-confidence data.
Extractions
0
document-intake extraction attempts on file
Pending Candidates
0
field suggestions still waiting for human review
AI Runs
0
document-intake AI runs logged for this claim
Usage Cost
$0.0000
estimated Claude cost for intake runs on this claim
Packet Generation
This creates the first extraction record and candidate suggestions. For now it uses artifact metadata, filename text, and notes as a deterministic placeholder until OCR and Claude-based intake are connected to this same route.
Ready to generate
0
Extraction Queue
Document Intake AI Runs
Claude Usage