Exporting Bulk Fhir

What it does

When your agent uses it

• The user needs population-scale note extraction from an EHR
• Data warehouse to feed OpenMed
• Mentions bulk export
• Cohort de-identification

Example prompts

• “/exporting-bulk-fhir”

Requirements

• Python 3

Workflow steps

7 steps, taken from the first numbered list in SKILL.md.

1. 1Obtain a SMART Backend Services token (system/DocumentReference.read, etc.).
2. 2Kickoff $export at the right level with _type (and _since for
3. 3Poll Content-Location until 200; read the manifest output[].
4. 4Download each NDJSON file (send the token if requiresAccessToken).
5. 5Stream each line → extract note text → openmed.deidentify →
6. 6Export findings to FHIR if needed (exporting-to-fhir,
7. 7DELETE the bulk job to free server storage.

What it can do on your machine

Read from SKILL.md and the folder at commit 34d7b8c. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Shell commands in SKILL.md call:

  • curl

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  Links to these hosts (documentation or services it may open):

  • hl7.org
  • github.com

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Exporting Bulk Fhir loads about 1.9k tokens when it runs. Until then it costs about 166 tokens; SKILL.md has 567 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~166

When it runs · the whole SKILL.md, loaded when a task matches

~1.9k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from maziyarpanahi/openmed at commit 34d7b8c, republished under its Apache-2.0 licence (© maziyarpanahi). 567 words, ~1,895 tokens.

name

exporting-bulk-fhir

description

Kick off and harvest a FHIR Bulk Data $export (system-, group-, or patient-level) and stream the resulting NDJSON into a batch OpenMed de-identification + NER pipeline at cohort scale. Covers the async kickoff (Prefer respond-async) -> poll Content-Location -> download NDJSON flow, the Bulk Data Access IG, _type/_since filters, and feeding DocumentReference/DiagnosticReport notes into openmed.deidentify in batch. Use when the user needs population-scale note extraction from an EHR or data warehouse to feed OpenMed, mentions bulk export, $export, NDJSON, Flat FHIR, or cohort de-identification. Pairs before the OpenMed de-id/NER pipeline.

license

Apache-2.0

metadata.project

OpenMed

metadata.category

fhir-interop

metadata.pairs

before

metadata.version

1.0

# 1) Kickoff (async). Ask for clinical-note-bearing resource types.
curl -s -X GET \
  'https://ehr.example/fhir/Group/cohort-42/$export?_type=DocumentReference,DiagnosticReport&_since=2024-01-01T00:00:00Z' \
  -H 'Authorization: Bearer <backend-services-token>' \
  -H 'Accept: application/fhir+json' \
  -H 'Prefer: respond-async' -D -
# -> 202 Accepted
#    Content-Location: https://ehr.example/fhir/bulkstatus/JOB123

# 2) Poll the status URL until complete
curl -s 'https://ehr.example/fhir/bulkstatus/JOB123' \
  -H 'Authorization: Bearer <token>'
# 202 + X-Progress while running; 200 + a manifest JSON when done:
# { "transactionTime": "...", "request": "...", "requiresAccessToken": true,
#   "output": [
#     { "type": "DocumentReference",
#       "url": "https://ehr.example/fhir/bulkfiles/dr-1.ndjson" },
#     { "type": "DiagnosticReport",
#       "url": "https://ehr.example/fhir/bulkfiles/dx-1.ndjson" } ] }

# 3) Download each NDJSON file (one FHIR resource per line)
curl -s 'https://ehr.example/fhir/bulkfiles/dr-1.ndjson' \
  -H 'Authorization: Bearer <token>' -o dr-1.ndjson

import base64, json, openmed

def note_text(resource: dict) -> str | None:
    # DocumentReference.content[].attachment.data (base64) or .url -> Binary
    for content in resource.get("content", []):
        att = content.get("attachment", {})
        if att.get("data"):
            return base64.b64decode(att["data"]).decode("utf-8", "replace")
    # DiagnosticReport.presentedForm[].data
    for form in resource.get("presentedForm", []):
        if form.get("data"):
            return base64.b64decode(form["data"]).decode("utf-8", "replace")
    return None

with open("dr-1.ndjson", "r", encoding="utf-8") as fh:
    for line in fh:                              # streaming, line by line
        resource = json.loads(line)
        text = note_text(resource)
        if not text:
            continue
        # De-identify every note before anything downstream sees it
        deid = openmed.deidentify(text, method="replace", policy="hipaa_safe_harbor")
        # Then NER on the de-identified text
        entities = openmed.analyze_text(
            deid.text, model_name="disease_detection_superclinical")
        # ... persist de-identified text + spans; never persist raw PHI

© maziyarpanahi, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/exporting-bulk-fhir of maziyarpanahi/openmed.

Open the folder on GitHubat commit 34d7b8c