📘 Context
A claims-processing AI assistant must enforce strict auto-approval limits and
fraud-flag escalation rules to prevent automated approval of suspicious or
out-of-scope claims that would otherwise require human adjuster review.
This policy acts as a deterministic gate before any auto-approval action,
ensuring fraud risk and policy-coverage edge cases get routed to humans.
🎯 The Goal
Write a claim auto-approval policy for an insurance AI assistant.
Variables you can use:
claim_type (Enum): "AUTO", "HEALTH", "PROPERTY", "LIABILITY", "WORKERS_COMP".
claim_amount_usd (Float): USD amount being claimed.
policy_coverage_limit_usd (Float): Coverage cap for this policy.
policyholder_tenure_years (Integer): Years policyholder has been with the company.
prior_claims_12mo (Integer): Number of claims in the last 12 months.
fraud_score (Integer): Fraud risk score 0-100.
documentation_complete (Boolean): Whether all required docs are submitted.
claim_within_policy_period (Boolean): Whether incident occurred during active coverage.
adjuster_review_requested (Boolean): Whether an adjuster has flagged for review.
Requirements:
- BLOCK if
claim_within_policy_period is False.
- BLOCK if
claim_amount_usd > policy_coverage_limit_usd.
- BLOCK if
fraud_score > 80.
- BLOCK if
documentation_complete is False.
- BLOCK if
claim_amount_usd > 5000 and adjuster_review_requested is False (auto-approval cap).
- ESCALATE if
prior_claims_12mo > 3.
- ESCALATE if
fraud_score > 50.
- ESCALATE if
policyholder_tenure_years < 1 and claim_amount_usd > 1000.
📚 Resources
- Start here:
docs/getting-started.md
- Check syntax:
quickstart/01_hello_world.csl
- See how examples are structured:
examples/community/
✅ Definition of Done
Submit a PR with: examples/community/insurance_claim_auto_approval.csl
📘 Context
A claims-processing AI assistant must enforce strict auto-approval limits and
fraud-flag escalation rules to prevent automated approval of suspicious or
out-of-scope claims that would otherwise require human adjuster review.
This policy acts as a deterministic gate before any auto-approval action,
ensuring fraud risk and policy-coverage edge cases get routed to humans.
🎯 The Goal
Write a claim auto-approval policy for an insurance AI assistant.
Variables you can use:
claim_type(Enum): "AUTO", "HEALTH", "PROPERTY", "LIABILITY", "WORKERS_COMP".claim_amount_usd(Float): USD amount being claimed.policy_coverage_limit_usd(Float): Coverage cap for this policy.policyholder_tenure_years(Integer): Years policyholder has been with the company.prior_claims_12mo(Integer): Number of claims in the last 12 months.fraud_score(Integer): Fraud risk score 0-100.documentation_complete(Boolean): Whether all required docs are submitted.claim_within_policy_period(Boolean): Whether incident occurred during active coverage.adjuster_review_requested(Boolean): Whether an adjuster has flagged for review.Requirements:
claim_within_policy_periodis False.claim_amount_usd > policy_coverage_limit_usd.fraud_score > 80.documentation_completeis False.claim_amount_usd > 5000andadjuster_review_requestedis False (auto-approval cap).prior_claims_12mo > 3.fraud_score > 50.policyholder_tenure_years < 1andclaim_amount_usd > 1000.📚 Resources
docs/getting-started.mdquickstart/01_hello_world.cslexamples/community/✅ Definition of Done
Submit a PR with:
examples/community/insurance_claim_auto_approval.csl