FRAME v1.0 · batch #1
FRAME / Batch #1 · May 28, 2026
FUNCTIONAL
12/24

Unified Health Record (UHR)

Aaqib-bashir1/unified-health-record

Principled consent-first design with best scope discipline in batch — but design phase only, no code, no FHIR/HL7, no AI layer. Blueprint not product.

CreatorAaqib Bashir
LicenseAGPL-3.0 with CLA
Stars★ 1
FHIRNo
Write opsNo
Stack Design only — no implementation
design-phasepatient-ownedconsent-firstappend-onlyAGPLCLAno-AIrecord-keepinglongitudinal-timeline
§ 01 / FRAME dimension scores eight axes · zero to three
PAI
2/3
Patient Agency
AI
1/3
Architecture Integrity
TM
0/3
Technical Maturity
SDP
2/3
Safety & Disclaimer
IS
1/3
Interoperability Stack
CA
2/3
CLAIM Alignment
SM
1/3
Sustainability Model
SH
3/3
Scope Honesty
§ 02 / Recommendation analyst-facing
MONITOR

Principled design, exceptional scope discipline, no code yet. Check back at spec-to-code phase. CLA governance structure worth noting in open-source health data governance framing.

§ 03 / Installation no runnable artifact

Not a runnable repository (spec, white paper, or design artifact). UHR is a design specification — README, governance docs, and an AGPL-3.0 + CLA license. There is no application to install yet. Read the spec to understand the consent-first append-only data model, then revisit once the project moves into the spec-to-code phase.

§ 04 / Full audit narrative + CLAIM

FRAME Audit: Unified Health Record (UHR)

URL: https://github.com/Aaqib-bashir1/unified-health-record
Date: 2026-05-28
Analyst: FRAME v1.0 / Synambix
Documents reviewed: README.md


One-Line Verdict

FUNCTIONAL | Score: 12/24 | Principled design with strong scope discipline — but design-phase only, no working code, no interoperability standards, and no AI layer. Scores at the FUNCTIONAL floor on the strength of its “what this is not” clarity alone.


Dimension Scores

DimensionScoreEvidence
Patient Agency Index2/3Patient owns data; consent-first access; explicit, revocable, scope-bound access model; correct framing. Design-phase only — no implementation to verify architecture backs up claims.
Architecture Integrity1/3Append-only data model and amendment-based corrections described in principles. No architecture doc or technical implementation visible. SECURITY.md exists but design only.
Technical Maturity0/3Explicitly “design and foundation phase. No production deployment exists yet.” Scope frozen, core models being defined. 1 star, 0 forks, no code visible beyond documentation.
Safety & Disclaimer Posture2/3”No AI diagnosis or treatment recommendations” explicit; “no automated clinical decision-making”; “context and clarity, not medical judgment.” No “not a medical device” disclaimer by name. SECURITY.md linked.
Interoperability Stack1/3No health interoperability standards mentioned. No FHIR, HL7, CCDA, or any EHR connector. “Manual digitisation supported” is the only data path mentioned. May be in the spec doc — not visible from README.
CLAIM Alignment2/3”Clarity over automation. Consent over convenience. Continuity over speed.” No AI in scope. Patient agency framing is correct. No explicit CLAIM literacy framework. No interrogative stance design.
Sustainability Model1/3AGPL-3.0 with CLA granting project lead right to relicense. Single maintainer. 1 star. Design phase. No community.
Scope Honesty3/3”What UHR Is NOT” section with explicit ❌ bullets. “Design and foundation phase” stated. “No production deployment exists yet.” The clearest scope statement of any project in the batch.
TOTAL12/24

Key Strengths

  • “What UHR Is NOT” is exemplary scope discipline. Six explicit exclusions: no AI diagnosis, no emergency monitoring, no billing/insurance, no predictive analytics, no automated clinical decision-making. Every patient AI project should write this section before writing code.
  • Append-only data model with amendment-based corrections. The data philosophy — nothing is overwritten, corrections are layered, conflicting data is preserved with attribution — is structurally sound and aligns with OwnChart’s immutability principle.
  • Consent model is correctly designed in concept. Explicit, time-bound, scope-bound, revocable access. If a patient is hidden history, doctors are informed the timeline is filtered. This is the right consent architecture for a personal health record.
  • AGPL-3.0 choice is principled. Any network deployment must release source. This is a strong copyleft choice that prevents commercial extraction without source disclosure.

Key Gaps

  • No code. This is the critical gap. All other dimensions are scored against principles that have no implementation. Principles without implementation are a hypothesis.
  • No interoperability standards. FHIR, CCDA, HL7 are not mentioned. Without a standards path, this system is a manual data entry tool — which may be intentional for the target geography, but limits the realistic patient population to those who can manually organize their own records.
  • CLA grants project lead right to relicense. This is a significant governance asymmetry. Contributors assign the right to sublicense, including to proprietary licenses. Combined with the “future commercial support” language, this could enable a bait-and-switch from AGPL to proprietary. This is worth watching.
  • No AI layer by design. The scope decision to exclude AI entirely is clean and honest. But it means UHR cannot compete with the other FOUNDATIONAL entries on any AI dimension. It is a data infrastructure play, not a patient reasoning platform.

CLAIM Assessment

Contextual Grounding: Not applicable — no AI layer by design. The system stores patient records but does not process them with AI.

Interrogative Stance: Not applicable by design. UHR provides “context and clarity, not medical judgment” — it is explicitly not a reasoning partner.

Associative Integration: Low. No interoperability standards means no path to connect this system’s records to other health data sources or AI tools without manual export.

Judgment Layer Activation: Correct in concept. The consent model, the explicit “no automated clinical decision-making” commitment, and the doctor-notification-when-history-is-filtered all put the patient in the judgment seat. No implementation to verify.

Methodological Transfer: Not in scope. This is a record-keeping tool, not a literacy platform.


Patient Agency Verdict

UHR starts from the right premise — information loss is the core problem, not lack of expertise — and designs toward the right solution: a patient-controlled longitudinal timeline with consent-first access. The principles are sound. The gap between principle and implementation is currently total.

The project scores at the FUNCTIONAL floor rather than ASPIRATIONAL for one reason: its scope discipline. It knows exactly what it is not, and it states this clearly. That discipline is harder to maintain than it looks, and most patient AI projects that score higher on technical maturity have lost it somewhere.

If this project produces working code that implements its stated design, it would score 15-17 — FUNCTIONAL with a clear path to FOUNDATIONAL if it adds FHIR import and an AI reasoning layer. Right now it is a well-designed blueprint.


Evidence Flags

  • CLA grants relicensing rights. Language: “Grants the Project Lead the right to sublicense or relicense contributions if needed in the future.” Combined with “future commercial support or alternative licensing models” language, this creates a governance risk for contributors. Not a current problem — a structural one to watch.

Recommendation

MONITOR. The design is principled and the scope discipline is exceptional. No code yet. Check back when the spec-to-code phase begins. The CLA governance structure is worth noting in any PRISM-N or open-source health data governance framing.