$ loading_
Checks claim-file readiness and compiles evidence for human review.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "adjudication-review" skill from askskill: 1. Download https://raw.githubusercontent.com/microsoft/aibast-agents-library/main/solutions/claims-processing/manual/skills/aibast_adjudication-review_02/SKILL.md 2. Save it as ~/.claude/skills/aibast_adjudication-review_02/SKILL.md 3. Reload skills and tell me it's ready
Review Jennifer Liu's theft claim file and identify what is still missing before evaluation. Separate observed evidence, inferred output, and recommended next steps.
A structured list of missing claim materials, separated into evidence, assessment, and next steps.
Perform a readiness review for the exact fictional claim record scope only; do not substitute another record. Summarize relevant support from policy terms, loss evidence, documents, and adjuster notes.
A source-backed readiness review showing which materials support the conclusion.
Based on the current claim file, explain what supporting materials are still needed and identify the authorized human review required before further handling.
A checklist of missing materials and the required human review, without any external actions.
Before evaluating a claim, a user can use it to check whether a specific claim file is review-ready. It considers policy terms, loss evidence, documents, and adjuster notes to identify gaps.
When a claim file needs to be escalated for human review, this skill can organize the source-backed evidence relevant to the request. Its output separates observed evidence, inferred output, and suggested next steps.
If claim handling is blocked by incomplete documentation, this skill can point out missing receipts, appraisals, or other support. It is suited for claim-file readiness questions in the synthetic pilot.
The README describes a claim-file readiness skill that combines policy terms, loss evidence, documents, and adjuster notes for human review. It emphasizes reviewing the exact fictional record requested, returning source-backed evidence, and clearly separating observed facts, inferred output, and next steps. It also requires explicit disclaimers that the result is not professional advice and that no approval, payment, filing, or other external action has occurred.
Combines policy terms, loss evidence, documents, and adjuster notes for human review.
Persona: Claims Adjuster
Prompt: What is missing from Jennifer Liu’s theft file before I can evaluate it?
Expected synthetic evidence: CLM-2025-7004, Receipts or appraisals.
It is used to answer whether a claim file is ready for review. According to the docs, it combines policy terms, loss evidence, documents, and adjuster notes to prepare evidence for human review.
Yes. The docs require separating observed evidence, calculated or heuristic output, and proposed next steps; they also require stating that the result is not legal, regulatory, insurance, lending, tax, investment, or financial advice.
No. The docs explicitly require stating that no approval, communication, filing, account change, payment, order, transaction, or other external action occurred, and that the required authorized human review must be named before further action.
Estimate policy-term settlement, coverage limits, and deductibles for claims processing.
Segment a fictional client portfolio and visualize health and churn indicators.
Triages fictional insurance claims and summarizes key intake details and priority.
Compare incentive options by synthetic cost, modeled lift, and net value.
Analyze client engagement signals across meetings, escalations, billing, and utilization.
Build fictional stakeholder maps and approval-gated executive engagement plans.
Review SIU indicators in claims and return explainable evidence with human review guidance.
Review contracts against playbooks with redlines, deviation flags, and impact analysis.
Prepare reviewable retention and service-recovery options for churn-related customer cases.
Prioritize churn reviews with transparent evidence and suggested next steps.
Checks legal AI outputs for orphan quotes and uncited claims.
Verify AI agent completion claims by checking output file existence and freshness.