HugoScore hugoscore.org

Full review

Health-GPT CAIHL draft report

Evidence-linked HugoScore draft report for a health AI tool that affects patients.

Agency checklist · October 30, 2026

Leans expands 8 of 12 yes

Document-based draft. Independent confirmation is incomplete. Hugo has not personally verified this service.

Assessed function: Omni Timeline LLC patient record-comprehension and visit-preparation app.

A no can mean missing documentation or institution-dependent control; it does not by itself establish harmful behavior. These categories are not clinical safety or product-quality ratings.

  • Public documents do not establish actual product behavior.
  • The independent review screened rules and selected evidence; it did not replicate all twelve answers for this tool.
See all twelve answers and their evidence
QuestionAnswerEvidence / reviewSource and reasoning
1. Told it is AIYesdocumented
Not independently replicated
With secure service providers necessary to operate the app, such as cloud hosting or AI processing, under strict data protection agreements. Source

Weak yes. The Privacy Policy refers to AI processing and 'AI models'. No in-product AI notice or consent step is documented, and the Terms of Use link (/terms-use) returns 404.

2. Operator and model maker namedNonot-documented
Not independently replicated
we're not affiliated with OpenAI Source

Omni Timeline LLC is named as operator and seller. Model maker is not named. The listing only says it is not affiliated with OpenAI.

3. Patient chooses to use itYesdocumented
Not independently replicated
Download Health-GPT and prepare for your next doctor visit with clearer information and better questions. Source

Consumer app downloaded and used by patients or caregivers.

4. Patient sets the taskYesdocumented
Not independently replicated
Ask questions about your medical records and get clear, plain-language answers Source

Patient chooses documents and questions and can mark entries private so the AI will not use them.

5. Patient can decline or stopYesdocumented
Not independently replicated
Download Health-GPT and prepare for your next doctor visit with clearer information and better questions. Source

Self-started: Q3 is yes and no care depends on using Health-GPT. Source is the Q3 source.

6. Patient sees the AI outputYesdocumented
Not independently replicated
Generated content, such as summaries, timelines, suggested questions Source

The generated timeline and questions are shown to the user.

7. Patient can correct or challengeNonot-documented
Not independently replicated
Quotation not recorded; evidence repair pending. Source

No route to correct or contest generated content is documented. User rights listed are access, delete and deletion assistance.

8. Patient can get a copyNonot-documented
Not independently replicated
Access your data Source

Privacy section 9 lists 'Access your data' with no export, download or copy mechanism. Play Store data safety: 'You can request that data be deleted' only.

9. Patient can deleteYesdocumented
Not independently replicated
You can delete your account, which will permanently delete all associated data. Source

Individual entries can be deleted in the app. Requests completed within 7 days. Technical logs kept up to 30 days.

10. Training or sale does not override patient choiceYesdocumented
Not independently replicated
We do not sell your data. We do not use your data for advertising. We do not train AI models on your personal data. Source

Same statements on the App Store and Play Store listings.

11. Patient can run it or take it elsewhereNonot-documented
Not independently replicated
Quotation not recorded; evidence repair pending. Source

Vendor-hosted app. No self-hosting or documented move-elsewhere route.

12. No advertising or data businessYesdocumented
Not independently replicated
We do not sell your data. We do not use your data for advertising. Source

Section 5: shares only with service providers necessary to operate the app, such as cloud hosting or AI processing. AI providers are not named. Play Store Data safety says the app 'may share' files and docs with third parties, and the App Store label says no data collected, both of which conflict with the policy. Policy used.

Initial coding: Claude Sonnet helpers, compiled by Claude Opus 5.5 (Anthropic). Selective review and corrections: GPT-6 Astra (OpenAI). Editor: Hugo Campos. Acceptance of corrections is not product verification.

Checklist rules · Independent review and limitations · Later evidence follow-up

Earlier report

Health-GPT: CAIHL reassessment

September 8, 2026. Published AI-assisted draft.

Health-GPT offers patient-directed record comprehension and preparation. Readable FAQ and privacy material support a qualified assessment. The terms could not be retrieved, and the earlier app-store discrepancy remains historical.

Scope and agency posture

Omni Timeline LLC's patient record-comprehension and visit-preparation app, not other similarly named GPT products. Preserve August 15 baseline 80.

Posture: Potentially agency-expanding, with vendor-opacity and disclosure caveats.

Axis: 80/100, retained provisionally. No numerical recalibration was performed.

Sources and documented findings

  • The current homepage and now-readable FAQ describe upload, timeline, multilingual questions and exclusion of user-marked private entries from answer generation. They frame the tool as preparation and information and say it is not designed for pregnancy or mental health.
  • After an initial retrieval failure, following the site's own link recovered the privacy policy. It states no sale, advertising or training on personal data, automated processing by service providers under agreements, content retention until deletion, up to 30 days for technical logs and account/data deletion within seven days. Provider names and correction/export rights are still absent.
  • The linked terms again returned 404. The August App Store privacy-label discrepancy was not newly checked and remains a historical unresolved finding.

Patient authority

Inference from the documented workflow: Patients select records, questions and excluded entries. The exclusion control is meaningful, but it controls answer context rather than proving that an upload was never processed or retained. Generated timeline correction remains unestablished.

Critical capacity

Editorial assessment: Visit-question preparation can support the patient's own reasoning and participation. A readable timeline must remain checkable against original documents, especially for dates and medications. Deleting an entry is not the same as correcting an extraction.

Informed control

Editorial assessment: Current policy statements support the existing qualified assessment but leave the processing parties and usable exit formats unclear. The newly readable FAQ resolves a retrieval gap without validating clinical behavior.

Assessment and limits

The patient-directed design and unresolved verification conditions remain consistent with the baseline. Keep the posture and number, refresh the policy evidence and stop treating the FAQ as unreadable.

Confidence: Low to medium for public design and policy, low for implementation.

Remaining uncertainty: Output correction, export, named processors, operative terms, current store label and extraction quality remain unresolved.

Published documentation is evidence of stated conditions, not proof of actual implementation. Unknowns did not receive automatic negative points. Funding, sponsorship, and public code do not determine agency by themselves.

Review provenance

  • Reviewer/model: OpenAI Codex / GPT-6.
  • Method: Focused public-source reassessment using CAIHL: patient authority, critical capacity, and informed control. Existing evidence plus one focused primary-source pass and at most one targeted follow-up. No live product testing. Numerical scores remain provisional editorial placements, not a new calculation.
  • Human review: Hugo Campos authorized publication of these AI-assisted draft reassessments on September 8, 2026. This does not claim comprehensive human verification of every finding.
  • Earlier review: 2026-08-15. Historical assessment. Earlier claims are not automatically reverified by this publication.