Patient-controlled health records AI
MyChart Takeout
MyChart Takeout is a free, source-public bookmarklet and local CLI by Josh Mandel that exports a patient's complete Epic/MyChart record into a single ZIP: problems, medications, lab and imaging results, visit notes, secure messages, C-CDA documents, implanted-device identifiers, and even the access log showing who viewed the record. The export runs entirely inside the patient's own signed-in browser tab. There is no server, no account, and no credential handling, so the privacy claim is architectural rather than contractual. The exporter is deterministic, with no language model in the data path, but the output is deliberately AI-ready through a consolidated PATIENT_SUMMARY.md, CSV indexes, and structured JSON the patient can hand to any AI tool they choose. Proxy support extends the export to children and dependents a caregiver manages.
Public-source research has been drafted, often with AI assistance. No comprehensive human review is recorded unless the profile provenance says so.
Summary judgment · 92% toward patient-directed
Strongly agency-expanding, with license-clarity, instance-variability, and access caveats
The tool moves a patient's complete Epic portal record into a local ZIP the patient fully controls, with no server in the data path and no language model in the export. The caveats are Epic-instance variability, an unstated software license, and the baseline requirement of portal access plus bookmarklet comfort.
Patient agency
How this tool changes agency
The export opens with a readable PATIENT_SUMMARY.md and CSV indexes rather than raw JSON alone, and the complete record supports visit preparation, second opinions, advocacy, archiving, caregiver coordination, and any downstream AI analysis the patient selects.
The tool copies records rather than generating judgments, and correcting source records still depends on the institution. But exporting the access log, complete notes, and messages gives patients durable evidence for challenging errors, disputed entries, or record-access questions.
Patient-facing signals
Who does this AI serve?
The tool serves patients, caregivers, and patient-developers who want their complete Epic record as local files. It has no institutional customer, no account system, and no monetization visible in public sources. It is not affiliated with or endorsed by Epic.
Can patients tell AI is involved?
The README states plainly that the exporter is deterministic and no language model is involved in the export itself. Agent-driven use of the CLI is described as an option the user chooses, not a hidden layer.
Can patients meaningfully choose?
Use is entirely voluntary and patient-initiated. The bookmarklet path asks only that a patient drag a button to the bookmarks bar, sign in as usual, and click Start, a far lower burden than token- or script-based alternatives. The deeper CDP/CLI mode remains available for technical users.
Can patients correct or challenge what the AI produces?
The tool copies records rather than generating judgments, and correcting source records still depends on the institution. But exporting the access log, complete notes, and messages gives patients durable evidence for challenging errors, disputed entries, or record-access questions.
Does it help patients understand or act?
The export opens with a readable PATIENT_SUMMARY.md and CSV indexes rather than raw JSON alone, and the complete record supports visit preparation, second opinions, advocacy, archiving, caregiver coordination, and any downstream AI analysis the patient selects.
Text findings
Who is left out or burdened?
Portal-access and digital-comfort burden
The tool only helps patients whose systems run Epic MyChart and who have an active portal account, a desktop-class browser, and comfort with bookmarklets or the DevTools console. The interface and outputs are English-only. Proxy export meaningfully includes caregivers of children and dependents, but patients without digital access or portal literacy are left out.
What happens to patient data?
Local-only by architecture
The export runs in the patient's browser tab and writes a ZIP directly to their downloads. The landing page is static, the bookmarklet is self-contained, and there is no backend to receive data. The tool never sees or stores passwords, using the session the patient already holds. In CLI mode, exports land only in local git-ignored directories. These claims are verifiable in the public source but have not been independently audited.
Are the clinical boundaries clear?
Clear that it is a data-access tool, not clinical advice
The tool retrieves and packages records. It does not diagnose, triage, or recommend treatment, and no clinical model runs anywhere in the path. The residual risk is downstream: what a patient or their chosen AI tool does with raw records is outside the tool's boundary.
Who defined what good looks like?
Maintainer-defined, source-auditable, with unusually strong self-verification
The maintainer, a widely known health-interoperability architect, defines quality through a public test suite of 110 tests, a synthetic-patient mock instance containing no PHI, and published parity provenance comparing implementations record-for-record. Public evidence shows no independent security audit, accessibility review, or patient-partnered evaluation, and the repository carries no explicit software license despite the open-source claim.
Review provenance
Criteria
Same HugoScore CAIHL-derived criteria for every tool; public criteria are displayed from site/data/criteria.json, with fuller method notes in SCORING_FRAMEWORK.md.
Reviewer
AI-assisted public-source draft prepared in Claude (Cowork); no named human reviewer recorded.
AI / model
Claude Fable 5 (claude-fable-5), first-pass draft.
Human review
No comprehensive human review has been completed or claimed for this draft profile.
Review date
2026-08-14
Limitations
No hands-on export with a real MyChart account, no security audit, no real-PHI code execution, no maintainer interview, no accessibility review, no legal determination. Claims are limited to the project's own public site, repository, and documentation.
Review method
Public-source review of the official landing page, GitHub repository README, package manifest, repository layout, output-format documentation, and parity-test provenance notes; no hands-on export with a real MyChart account, security audit, real-PHI code execution, maintainer interview, accessibility review, or legal review.
Draft profile · Medium draft, source-public, no independent audit