Last year's checkup wrote A1c, this year's panel HbA1c, the new clinic Glycated Hemoglobin. One test, three names, nothing to compare. Mirobody takes health information from any source, in any format, under any name, and settles it into one language and one system, then answers over that record, every number citing its source: traceable, comparable, chartable. How has my blood pressure moved? Are mom's diabetes markers improving? What changed across my child's checkups? Self-host it all, and your health record stays in your hands.

Three files, three names for the same test, one standard code. The agent finds all three, aggregates the trend, and names the file every number came from.

What Mirobody does

• One record for the whole family. Invite a partner, a parent, even a child

• Every source, one record. Garmin, Oura and Whoop connect directly;

• Say how you feel, in your own words. Type headache since last night,

• No hallucinations, everything traceable. Every indicator lands in one

• The agent reasons only over coded data. Trends by minute, hour, day,

• Genotypes as facts, not verdicts. Upload a 23andMe, AncestryDNA,

• Runs on a laptop. Three containers: Postgres, the server and the worker.

• Your model, your key, your data. Model calls go to the model you chose.

Try it in 60 seconds

One command, five spellings: watch which ones it recognises, and which one it refuses. No key, no config, no network, and with uvx, no install either:

uvx mirobody resolve "LDL cholesterol" 血红蛋白 ヘモグロビン "空腹血糖(GLU)" 血脂

血红蛋白 and ヘモグロビン: two languages, one code, 718-7. 血脂 (lipids) names a category, not one observation, so it resolves to nothing. The resolver would rather return nothing than guess a code, because a wrong one puts two different tests on the same trend line.

from mirobody.engine import resolve, resolve_reading
resolve("血红蛋白").loinc                                # '718-7'   any language, one code
resolve("total cholesterol").loinc                     # '2093-3'  [Mass/volume]
resolve_reading("total cholesterol", "5.0", "mmol/L").loinc  # '14647-2' [Moles/volume]
resolve_reading("total cholesterol", "193", "mg/dL").loinc   # '2093-3'  the unit picks the code
resolve("中性粒细胞百分比").loinc                          # '26511-6' Neutrophils/Leukocytes
resolve_reading("中性粒细胞", "62 %", None).loinc          # '26511-6' a percentage...
resolve_reading("中性粒细胞", "4.2", "10*9/L").loinc       # '26499-4' ...and a count are two codes
resolve("血脂").resolved                                 # False    a category, not an observation
from mirobody import standardize_reading                # the same answer as a FHIR Observation
standardize_reading("血红蛋白", "13.5", "g/dL")["code"]["coding"][0]["code"]  # '718-7'

Pass the value and the unit when you have them. A different unit means a different test, and LOINC folds that into the code's own identity, so one name is deliberately several codes. → Engine reference · Indicators

Collect · Translate · Agent

An indicator takes three steps from arriving to being cited. Each one leaves a trace, so the answer at the end can be followed back to the page it came off:

| Stage | What it does | Where | | --- | --- | --- | | ① Collect | Lab reports, wearables, phone photos, genetic files, all pulled in. The source file is kept as it was, so every indicator points back to the page it was read from. | collect/ | | ② Translate | One name to one code, one unit to UCUM, offline and deterministic. A1c, HbA1c and Glycated Hemoglobin become the same test here, and 头疼 and headache the same complaint (ICPC-3). | engine/ · translate/ | | ③ Agent | Ask over the coded record. Trend a value by minute, hour, day, week or month; get count, min, max, avg or change over any window in one call; compare across labs and devices, because they share one code. It charts the result in its reply, reads medications and genetic variants too, and names the file every number came from. | agent/ |

① records how the source spelled it, ② decides what it actually is, ③ answers on that footing. Comparing a number across two labs, charting three years of it, computing a baseline: all of it rests on the code ② hands over.

The agent does not have to be ours. Every tool it uses is served at /mcp as well, gated per user. Claude Desktop, Cursor or your own loop run the same tools over the same record, and get back the same indicators. The vocabularies need no server at all: uvx mirobody mcp serves them over stdio, with no database and no key, to any MCP client.

Garmin, Oura and Whoop connect with your own credentials from each vendor; the setup guide walks it through. Apple Health goes another way: a client on the phone hands the data over, so any band, ring or scale reaches your record the moment it writes into Apple Health, with nothing to integrate here at all.

Privacy

Nothing leaves your machine except calls to the model you chose, and to a device vendor once you link one. Reading a photo of a report, pulling indicators out of a PDF, splitting a sentence you typed into the journal, answering your question: all four call the model. Which provider and which model is the one key in your .env. A linked Garmin, Oura or Whoop is called through its own API, for what it recorded and nothing else.

② Translate stays local entirely: a name to a code, a unit to UCUM, looked up against a bundle that ships inside the package. No key, no network, no GPU, no model. Your record lives in your own Postgres, in containers you run, and nothing here reports usage anywhere.

One model key to bring yourself. deploy.sh generates the database, signing and encryption secrets in .env. Put an OpenRouter key (OPENROUTER_API_KEY), a Gemini key (GOOGLE_API_KEY), an OpenAI key (OPENAI_API_KEY) or an Anthropic key (ANTHROPIC_API_KEY) in the .env beside compose.yaml, then docker compose restart. DeepSeek, DashScope or any OpenAI-compatible gateway works alone too. Which model chats, which reads report photos, which extracts indicators and which embeds are four lines in config.llm.yaml, and that file names the variable (api_key: OPENROUTER_API_KEY), never the secret. mirobody doctor prints what each surface selected, and names the fix where one has nothing.

The repository's config shows placeholders; deploy.sh replaces them for the container stack. Encryption at rest does not yet cover every field. Before this reaches a network you do not control, read SECURITY.md: it also lists exactly what the server calls off your machine.

🚀 See it end to end

git clone --depth 1 https://github.com/thetahealth/mirobody.git && cd mirobody
./deploy.sh                       # pulls the app image; starts Postgres, server and worker → http://localhost:18060

(--depth 1 skips the history of superseded frontend builds; drop it if you plan to send a pull request.)

deploy.sh creates local secrets in .env and pulls thetahealth/mirobody:1.5.3 from Docker Hub. The image already carries the terminology bundle, so Docker users do not need Git LFS. When the daemon cannot reach Docker Hub it uses the docker.1ms.run mirror, and when the image cannot be pulled at all (a branch, or a release not yet published) it builds it from the checkout, which then needs git lfs pull. Upgrading a 1.5.2 stack: docs/backup-restore.md. For a second checkout, set COMPOSE_PROJECT_NAME and host ports in its .env. A Docker daemon that rejects named volumes can use compose.override.yaml.example for bind mounts.

Sign in on the Email code tab as [email protected], code 111111, no mail provider needed. An account of your own is one request away:

curl -X POST localhost:18060/password/register -H 'Content-Type: application/json' \
     -d '{"email":"[email protected]","password":"at-least-8-chars"}'

SEED_DEMO_DATA is on by default, so two accounts are already there with 2,019 indicators between them: you, and [email protected], who shares her record with you view-only. Set it to false to hold real data and neither account is created. Settings → Add member covers someone who will never sign in at all, a parent, a child, with a record you hold on their behalf.

Drop a file on the Data page and watch it become indicators. demo/upload/ holds four files the seed deliberately leaves out: a lab PDF, a phone photo of a printed report, a spreadsheet and another lab's CSV export. Each analyte comes out with a value, a unit and a LOINC code, linked back to the page it was read from.

Ask how the cholesterol has moved and the agent finds every file that carries it: one lab writes Cholesterol, Total where the others write Total Cholesterol-TC, and both are 14647-2. It charts the trend and names the file each number came off: 4.60 → 4.45 → 4.38 mmol/L. Ask for a baseline or a monthly average instead and the same tool aggregates over the whole record, rather than handing back rows for the model to add up itself.

Ask the same question of the record shared with you and it is a different person's answer, from data you can only view. That sharing is a **care circle**: invite-only, off by default, and strictly permission-checked.

Each of those three words is one check, and they all live in one function. resolve_subject is the only way an account reaches a record that is not its own — being in a circle together grants nothing by itself.

→ The four-minute walkthrough · examples/06_care_circle_rules.py prints the whole sharing decision table offline · Docker deployment · Configuration

Check any of it yourself

Every figure below comes with its source: a command you can run, or a public dataset.

• 296/296 on the tests an ordinary checkup prints, in English, Chinese

• 13 wearable vendors, read field by field: 289 of 447 fields carry a LOINC

• One genotype truth in ten file shapes: 13 public 1000 Genomes calls

• Coding decisions you can replay: 13 synthetic readings and 26

• Three open benchmarks, public datasets, one command each: longitudinal

• The package names the vocabulary that answered you:

• 316 standard device indicators and 328 UCUM units with dimensional

• pip install mirobody is 2 packages, numpy the only dependency.

🔌 Use it, extend it

| You want | Do this | | --- | --- | | Offline resolution and units in your code | pip install mirobody — no key, no network | | A document turned into indicators | pip install 'mirobody[parse]' — PDF, image, Excel, Word, PowerPoint, text; only a scanned page reaches a vision model | | These tools in Claude Desktop, Cursor or your own loop | Settings → MCP: every agent tool is also served at /mcp, gated per user | | Coding in any MCP client, no server | uvx mirobody mcp (stdio): readings to FHIR Observations with their code, complaints to ICPC-3, units; no key, no database | | Your app talking to a deployment | The HTTP API, against the deployment you run — your app, your data layer | | A new tool or device provider | Drop a file into mirobody/agent/tools/ or mirobody/collect/providers/ and restart, or pip install a package declaring a mirobody.providers / mirobody.tools / mirobody.agents entry point | | Your own agent harness | pip install 'mirobody[agent]' for the middleware and virtual-filesystem backends, or point AGENT_DIRS at your directory to replace the shipped agent outright |

→ API overview · MCP integration · Adding tools · Bringing your own agent

🤝 Contributing

The highest-leverage contribution is a term the resolver gets wrong. Run mirobody resolve ""; if the answer is wrong or empty, report it or add a row to resolver_overrides.tsv plus a case to test_engine_coverage.py — the coverage score is the review.

pip install -e '.[test]' && pytest -q && lint-imports

→ CONTRIBUTING.md · Local Python setup · Repository layout · Roadmap · CHANGELOG · SECURITY

📚 Documentation, and what shaped this

docs.mirobody.ai, in English and Chinese — start at the Quickstart or the API reference. The Quickstart also ships with the code, as docs/quickstart.md, so it cannot drift from the commands in this repository; the contributor guides are in docs/.

Mirobody's design draws on the following standards and projects, with thanks: HL7 FHIR, Regenstrief Institute (LOINC), UCUM, OHDSI OMOP, Open Wearables, Open mHealth / IEEE 1752, wearipedia, dlt / Airbyte / Singer, deepagents and LangChain. The terminology licences this ships under are in LICENSE-3RD-PARTY.