Sample · synthetic data · what you'll receive

This is what an Evaluation Report looks like.

Below is the full Evaluation Report we'd ship for a realistic Bremerton, WA single-family FHA case. The names, addresses, and hashes are synthetic, but the structure, the findings detail, and the framing are exactly what your real customers will receive.

Before you scroll

The report below is long — about three printed pages. Most of what makes it useful happens in four places. If you're skimming, these are the ones to read:

  • Section 4 (Findings detail). The actual catches: one critical finding (lead-paint missing on a pre-1978 property) plus two material ones. Each has the citation, plain-English message, and an improvement recommendation.
  • Section 5 (Improvement guidance, ranked). The consolidated to-do list ordered by what would block your deal versus what would slow it down.
  • Section 6 (How to re-verify offline). The exact CLI command anyone (regulator, attorney, carrier) can run without our service. Two files in, six lines out.
  • Section 7 (What this Record attests — and what it does not). The framing that keeps you safe. Six things the seal does prove, six things it deliberately doesn't.
Athena Systems North Star

Evaluation Report

Decision Record id decrec_2026-04-15T14-22-08Z_a4f2c1e8
Issued 2026-04-15 14:22:08 UTC

1. Subject and assignment

Property address 1247 Elm Street, Bremerton, WA 98310
Effective date 2026-04-15
Loan program FHA-insured Title II forward — Section 203(b) (FHA)
Appraiser Sample Appraiser, SR-EXAMPLE-001-WA
Client / intended user Sample Lender Trust
Report form URAR-1004
Submission id sub_2026-04-15T14-21-50Z_bremerton_elm

Single-family residence built 1962, owner-occupied, full interior + exterior inspection performed 2026-04-12. Subject is pre-1978 construction, triggering FHA lead-based-paint and Title X disclosure considerations.

2. Decision Record at a glance

This evaluation is a sealed event. The block below records the exact bytes evaluated, the exact rules applied, and the exact moment the evaluation occurred. Anyone holding the North Star public key can independently re-derive the seal and confirm the record has not been altered.

Work file content hash sha256:7c1f6a8b9d4e2c5f1a3b8c9d4e2f1a3b8c9d4e2f...
Rule set applied us-multi.appraisal.2026-Q4-approved (360 rules)
Engine version 0.1.0
Findings hash sha256:5e3a9c2b7f1d4a8c5e3a9c2b7f1d4a8c5e3a9c2b...
Evaluation timestamp 2026-04-15 14:22:08 UTC (RFC 3161 TSA: timestamp.digicert.com)
Chain predecessor sha256:c8d4e2f1a3b8c9d4e2f1a3b8c9d4e2f1a3b8c9d4...
Signing key id north-star-prod-2026-Q2
Algorithm Ed25519 over canonical CBOR

How to verify offline: see Section 6.

3. Findings summary

SeverityCount
Critical — loan-eligibility blockers; the deal cannot proceed until cured 1
Material — report defects requiring correction before sign-off 2
Advisory — best-practice nudges; non-blocking 1
Compliant — rules that fired and passed 38
Total rules evaluated 42
Rules not applicable — applicability gate didn't match this submission 317
Action required. This submission has critical-severity findings. Resolve them before delivering the report to the lender.

4. Findings detail

Critical FED-FHA-4000-1-II-D-5-LEAD-PAINT

Lead-based-paint assessment present on pre-1978 property

HUD-4000-1 Section II.D.5 — source

Subject was built in 1962. FHA Handbook 4000.1, Section II.D.5 requires a lead-based-paint assessment for any property constructed before 1978. The submitted work file does not record a value for content.subject_property.lead_paint_assessment. The deal cannot proceed to closing until either (a) the appraiser certifies no defective paint, or (b) defective paint is documented with a remediation plan.

Improvement guidance. Add the lead-paint assessment to the URAR addendum. Use one of the three accepted values: no_defective_paint (no chipping / peeling observed on a building component); defective_paint_with_repair_plan (observed defects + remediation scope documented and bid); defective_paint_undocumented (observed but no plan — will block FHA insurance). Photograph any observed defects.

Material FED-FHA-4000-1-II-D-3-COMPARABLE-RECENCY

Closed comparables are within 12 months of effective date (or older comps are explained)

HUD-4000-1 Section II.D.3 — source

Comparable Sale #3 (953 Pine Avenue) closed 2024-12-08, more than 16 months before the effective date. The work file's comparable_sales_within_12_months boolean is false and no per-comp age explanation is recorded. FHA Section II.D.3 expects closed comparables within 12 months; older comps are permitted only when the addendum justifies the choice (e.g., rural / low-volume market).

Improvement guidance. Either substitute Comp #3 with a closer-in-time sale (preferable — the West Bremerton submarket has had adequate transaction volume in the trailing 12 months), or add a per-comp narrative explaining the selection: why this older comp is more locationally / physically similar than any alternative within the 12-month window. The narrative goes in the URAR addendum and the work-file's per-comp age_explanation field is then populated.

Material FED-FNMA-SG-B4-1-2-01-FLOOR-PLAN-SKETCH

Report includes a floor plan or building sketch with GLA calculations

FNMA-SELLING-GUIDE Section B4-1.2-01 — source

The submitted work file includes a floor-plan sketch but the content.report.gla_calculation_present flag is false. Fannie Mae B4-1.2-01 requires both the sketch AND the supporting GLA calculations (typically presented as a per-area dimension table next to the sketch, summing to the reported GLA figure).

Improvement guidance. Add the GLA calculation as a supplementary exhibit alongside the existing floor-plan sketch. Show dimensions per room or per finished area, sum to the reported gross living area, and confirm the sum matches the URAR's GLA field (1,847 sq ft as reported on this submission). The exhibit can be hand-drawn on the sketch margin or auto-generated by your floor-plan software.

Advisory FED-FNMA-SG-B4-1-3-09-MARKET-CONDITION-SUPPORT

Market-condition adjustments are supported by trend data

FNMA-SELLING-GUIDE Section B4-1.3-09 — source

Comparable Sales #2 and #3 carry positive market-conditions adjustments of $4,500 and $9,200 respectively. The work file's market_condition_adjustment_supported boolean is true (attested), but the supporting trend data is not referenced in the report. (advisory; consider but not required for compliance.)

Rules that passed

The 38 rules below fired against this submission and produced no findings. We show eight representative passes; the remaining thirty are listed in the full Decision Record JSON.

  • WA-RCW-18.140-050-A — Appraiser must hold an active certification
  • WA-WAC-308-125-200-USPAP-EDITION — USPAP edition referenced
  • USPAP-SR1-4-COMPS-MIN — At least three closed comparables
  • FED-FHA-4000-1-II-D-1-ROSTER-ELIGIBILITY — Appraiser on the FHA roster
  • FED-FHA-4000-1-II-D-3-PHOTO-FRONT-EXTERIOR — Front-exterior photograph included
  • FED-FHA-4000-1-II-D-4-FEMA-FLOOD-ZONE — FEMA flood zone identified (Zone X — minimal risk)
  • FED-FHA-4000-1-II-D-5-FOUNDATION-STRUCTURAL — Foundation / structural elements assessed
  • FED-FHA-4000-1-II-D-6-FHA-CERTIFICATION — FHA-specific certification present
  • … and 30 more.

5. Improvement guidance — ranked

The recommendations below are the consolidated, ranked list of actions to take before resubmitting this work file. The order follows severity (critical → material → advisory) and within severity, chronological order in your workflow.

  1. Add lead-based-paint assessment to the URAR addendum critical

    Pre-1978 construction triggers FHA II.D.5. Select one of: no defective paint observed; defective paint documented with repair plan; defective paint without repair plan (this third option blocks FHA insurance). Photograph any observed defects.

    Driven by: FED-FHA-4000-1-II-D-5-LEAD-PAINT.

  2. Replace Comp #3 or add an age-justification narrative material

    Comparable Sale #3 is 16 months old. Either swap in a sale closed within the last 12 months from the West Bremerton submarket, or document in the addendum why this older sale is more similar than alternatives.

    Driven by: FED-FHA-4000-1-II-D-3-COMPARABLE-RECENCY.

  3. Add GLA calculation exhibit alongside the floor-plan sketch material

    Per-area dimensions summing to the reported 1,847 sq ft GLA. Can be added to the existing sketch as marginal annotation or as a separate exhibit page.

    Driven by: FED-FNMA-SG-B4-1-2-01-FLOOR-PLAN-SKETCH.

  4. Reference the supporting trend data for market-conditions adjustments advisory

    Comps #2 and #3 carry positive market-conditions adjustments. Consider citing the West Bremerton 12-month rolling price trend (Form 1004MC or equivalent) in the addendum so the adjustments are inspector-defensible.

    Driven by: FED-FNMA-SG-B4-1-3-09-MARKET-CONDITION-SUPPORT.

6. How to re-verify this Decision Record offline

This Record can be re-verified at any time, without our service running, using only the Record file and our public key. The verification confirms that the work file content, the rule set, the timestamp, and the findings have not been altered since seal time.

Files you need:

  • decrec_2026-04-15T14-22-08Z_a4f2c1e8.json — the canonical Decision Record (in this download)
  • north-star-public-key-north-star-prod-2026-Q2.pub — the North Star public-key file (verify.athenanorthstar.com/keys)

One-line check (Bash):

athena verify --record decrec_2026-04-15T14-22-08Z_a4f2c1e8.json \
              --public-key north-star-public-key-north-star-prod-2026-Q2.pub

Expected output:

record_hash      OK  (matches canonical bytes)
signature        OK  (Ed25519 over canonical CBOR)
timestamp        OK  (RFC 3161, timestamp.digicert.com)
chain            OK  (predecessor hash matches)
findings_hash    OK  (derived from canonical findings)
content_hash     OK  (derived from canonical work file submitted)
=> verified: true

The verifier is a small, focused, open-source tool published by Athena Systems. It runs locally with no Athena Systems service required — you only need the Decision Record and the public key. Anyone (regulator, auditor, opposing counsel) can install and run it independently, and can read the verifier's source code before trusting its output.

7. What this Decision Record attests and what it does not

This section is the legal-defensibility framing. Read it carefully. The Decision Record is an evaluation event witness, not a content attestation.

What this Decision Record attests

  • The exact bytes evaluated. The SHA-256 hash of the canonical work file submitted at evaluation time matches the content_hash field above.
  • The exact moment of evaluation. The RFC 3161 timestamp (timestamp.digicert.com) places the event at 2026-04-15 14:22:08 UTC. Counter-signed by an independent TSA; not back-datable.
  • The exact rule set applied. The 360 rules in force at evaluation time are pinned by the us-multi.appraisal.2026-Q4-approved hash. Any change to the rules produces a different hash.
  • The exact findings produced. The findings_hash is deterministic; anyone re- running the same engine on the same content with the same rule set produces the same hash.
  • The signing identity. Ed25519 signature ties the Record to the north-star-prod-2026-Q2 signing key.
  • Continuity of evidence. The predecessor_hash links this Record into the hash chain. Flipping a byte of any prior Record breaks the chain at the next verify.

What this Decision Record does not attest

  • That the work file is accurate. The Record witnesses what was submitted. We did not source-verify the comparable data, the condition, the income, or any other content. Garbage in, garbage out.
  • That the appraiser's underlying valuation is correct. Our rules evaluate structural compliance, not the value conclusion. A passing evaluation does not mean a defensible valuation.
  • That observed condition matches reported. If a report says C3 and the property is C5, the rule still passes. The Record witnesses what was reported, not what was physically observed.
  • That comparable-sale source data is accurate at source. MLS, deeds, prior-sale histories are taken from the work file as submitted. The Record does not warrant upstream sources.
  • That the report passes any standard outside our rule set. A passing evaluation reflects only the specific rules in force at evaluation time.
  • That the evaluation is legal advice or professional opinion. North Star is a software tool. Use it accordingly.

The plain-English summary

You submitted a work file. We ran our published rules against it on the date and time shown. The cryptographic seal proves we did exactly that, with exactly that content, against exactly those rules. It is a tamper-evident record of the evaluation event. It is not — and must not be presented as — a third-party attestation that the appraisal itself is accurate.

8. Rule set provenance

Rule set idus-multi.appraisal.2026-Q4-approved
Effective from2026-10-01
Approved byJustin Moriarty (appraisal SME, licensed appraiser WA + OR) + Darin Molnar (CTO, final tech say)
Approval date2026-05-29
Rule-set hashsha256:b366cca8e2f1a3b8c9d4e2f1a3b8c9d4e2f1a3b8…

The full rule set is version-controlled internally and pinned by hash. The hash above identifies the exact rule definitions in force at evaluation time — any change to any rule would produce a different hash. The rule set is inspectable on request by regulators, auditors, and customers under the customer agreement.

9. Engine + tooling provenance

Engine version0.1.0
HashingSHA-256 (FIPS 180-4)
SignatureEd25519 (RFC 8032)
TimestampingRFC 3161 (timestamp.digicert.com)
Canonical formDeterministic CBOR (RFC 8949)
ReproducibilityVerified by golden-file CI on every engine release

Athena Systems North Star — Evaluation Report v1.0. This is a software-generated report. The cryptographic Decision Record at decrec_2026-04-15T14-22-08Z_a4f2c1e8.json is the authoritative artifact; this document is a human-readable summary of that Record.

What you just read

That's the entire deliverable for one evaluation. A working appraiser receives one of these every time they run a work file through North Star — in seconds, with the cryptographic record attached, ready to archive or hand to a reviewer.

The structure is the same on every report:

  • Sections 1–3 give the assignment context and the seal's cryptographic fingerprint.
  • Section 4 is the day-to-day value: findings with citations and per-finding improvement steps.
  • Section 5 ranks the improvements so you know what to fix first.
  • Sections 6–7 are the long-tail defensibility surface: how to re-verify, and what the seal deliberately does and does not claim.
  • Sections 8–9 document the rule set and engine versions, traceable to rule-set hashes.

Want one of these for your next assignment?

North Star is rolling out to Washington and Oregon appraisers first. Join the waitlist and we'll reach out when your state opens.

Join the waitlist