Open readiness standard

INSURANCE.md is the README for insurance readiness.

It turns scattered company risk facts, contracts, controls, policies, claims history, and evidence into a portable packet an agent can maintain, a human can review, and the business can own.

What it is

INSURANCE.md is a structured readiness file for companies preparing for insurance review, applications, renewals, due diligence, or contractual requirements.

  • company-owned instead of trapped in inboxes or third-party portals
  • portable across internal teams, advisors, reviewers, and diligence workflows
  • evidence-linked instead of assertion-only
  • diff-friendly across renewal periods
  • safe for agents to maintain with human review

What it is not

INSURANCE.md is not a policy, coverage indication, warranty, binder, or replacement for official applications. It should not make unsupported insurance advice claims or submit private data without human authorization.

Why this changes the power balance.

Companies should not have to guess what matters, repeat the same facts every renewal, or depend on one external party's memory. The packet makes risk context portable enough for agents to reason over and structured enough for humans to challenge.

Own the source of truth

A practical advantage for founders, CFOs, legal, security, operations, and the agents working on their behalf.

Ask better questions

A practical advantage for founders, CFOs, legal, security, operations, and the agents working on their behalf.

Compare advice against evidence

A practical advantage for founders, CFOs, legal, security, operations, and the agents working on their behalf.

Reuse work across renewals

A practical advantage for founders, CFOs, legal, security, operations, and the agents working on their behalf.

Prepare before pressure

A practical advantage for founders, CFOs, legal, security, operations, and the agents working on their behalf.

Keep authorization explicit

A practical advantage for founders, CFOs, legal, security, operations, and the agents working on their behalf.

V0 packet structure

The initial standard should be stable enough for agents to use, but simple enough for companies to adopt without a platform.

Company profile

Operations and revenue

Products and services

AI, model, and data usage

Contracts and insurance requirements

Current policies

Controls and evidence

Claims and circumstances

Gaps and unknowns

Human attestation

example/INSURANCE.md
---
schema_version: "0.1"
company: "Acme Agents, Inc."
generated_at: "2026-06-01"
prepared_by: "company_agent"
readiness_state: "review_required"
confidence: "medium"
---

# INSURANCE.md

## Business Summary
Acme Agents builds operational AI agents for finance and security teams.

## Insurance Requirements
- Enterprise customer contract requires Tech E&O, cyber, CGL, and D&O.
- Cyber minimum: $2M aggregate.
- Contract asks for notice of cancellation within 30 days.

## Evidence Index
- SOC 2 Type II report: evidence/security/soc2-2026.pdf
- Customer MSA: contracts/customer-msa-acme-bank.pdf
- Current cyber policy: insurance/cyber-policy-2025.pdf

## Gaps
- Confirm revenue split by geography.
- Confirm whether any model outputs are used for regulated decisions.
- Confirm known claims, incidents, or circumstances.

## Human Review
Status: not attested
Reviewer: pending

From artifact to protocol.

The goal is a standard agents can generate, validate, diff, and export. Clara is the steward of the format, benchmark, skills, workspace, and intelligence layer around it.

CapabilityPublic standardClara platform
GenerateAgent skills create a packet in the company workspace.Hosted packet workspace and guided review.
ValidateLint structure, gaps, risky claims, and evidence links.Saved validation history and benchmark reports.
ExportExecutive summary, evidence index, review checklist.Application-support exports, hosted review, and workflow routing.