OpenMed is local-first healthcare AI software that extracts clinical information and removes personally identifying details from clinical text on hardware controlled by the user. Healthcare developers use its Python runtime, Apple Silicon and mobile SDKs, and browser support for on-device clinical NER and PII de-identification.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add maziyarpanahi/openmed --skill auditing-deidentification-runsgit clone --depth 1 https://github.com/maziyarpanahi/openmedWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/maziyarpanahi/openmed/auditing-deidentification-runs)<a href="https://agentmods.dev/skills/maziyarpanahi/openmed/auditing-deidentification-runs"><img src="https://agentmods.dev/badge/skills/maziyarpanahi/openmed/auditing-deidentification-runs/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/maziyarpanahi/openmed/auditing-deidentification-runs"><img src="https://agentmods.dev/badge/skills/maziyarpanahi/openmed/auditing-deidentification-runs.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00155 | $0.01801 |
| Opus 5 | $0.00077 | $0.00901 |
| Sonnet 5 | $0.00031 | $0.00360 |
| Haiku 4.5 | $0.00015 | $0.00180 |
Grade A, and why
auditing-deidentification-runs scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 142 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Auditing de-identification runs
deidentify(..., audit=True) returns an AuditReport instead of the rewritten
text: a deterministic, PHI-free record of every redaction decision —
offsets, label, detector confidence and threshold, the action taken, content
hashes, model provenance, and projected residual risk. Sign it to make it
tamper-evident, retain it as compliance evidence, and verify it later without
ever touching the original text. Runs on-device.
When to use this skill
Use it whenever a de-identification needs to be defensible: regulatory retention, internal review, reproducibility checks, or proving to an auditor that identifiers were detected and handled — all without storing or exposing the PHI itself.
Quick start
import openmed
note = "Patient John Doe (MRN 1234567) seen 2024-03-02. SSN 123-45-6789."
# Returns an AuditReport, NOT a DeidentificationResult, when audit=True.
report = openmed.deidentify(note, policy="hipaa_safe_harbor", audit=True)
# Make it tamper-evident with a release HMAC key (keep the key in a vault).
report.sign(b"my-release-hmac-key", key_id="release-2026")
# Persist the no-PHI report (no plaintext identifiers inside).
import json
with open("deid_audit.json", "w") as fh:
json.dump(report.to_dict(), fh, indent=2)
# Later: verify integrity. Optionally bind to the exact texts via their hashes.
ok = report.verify(b"my-release-hmac-key", original_text=note)
assert ok
What an AuditReport contains
AuditReport (from openmed.core.audit) carries no plaintext PHI. Key fields:
| Field | Meaning |
|---|---|
policy |
policy profile name in effect |
resolved_profile |
the concrete settings applied (method, model, thresholds, language, sweep flags) |
detectors |
list[DetectorInfo] — provenance of each detector |
safety_sweep |
structured-ID sweep metadata (patterns version, source) |
spans |
list[AuditSpan] — one entry per detected identifier |
thresholds |
per-label confidence thresholds used |
residual_risk |
projected leakage / re-identification risk summary |
openmed_version |
library version that produced the report |
manifest_hash |
hash of the model manifest used |
document_length |
character length of the input |
input_hash |
sha256: hash of the original text |
deidentified_text_hash |
sha256: hash of the de-identified output |
repro_hash |
deterministic hash over the canonical payload |
signature |
AuditSignature once .sign() is called (else None) |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 142 lines · 155 tokens per session scan A b716ac4201cf
auditing-deidentification-runs is a skill published in the GitHub repository maziyarpanahi/openmed (5,290 stars, last pushed today), licensed Apache-2.0. It adds 155 tokens to every session and 1,801 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
benchmarking
Use this skill when the user wants to benchmark an MLX-VLM change and present the numbers in a PR — fork-vs-main A/B comparisons, isolated-module micro-benchmarks, median-of-N timing with warmup, peak-memory reporting, correctness checks, parameter sweeps, and self-contained reproducible bench scripts to paste into a…
convert-quantize
Use this skill when the user wants to convert a Hugging Face model to MLX or quantize/dequantize one with mlxvlm.convert, including bits and group size, quant modes (affine, mxfp4, nvfp4, mxfp8), RTN vs AWQ, mixed-bit recipes, dtype casts, calibration (text or multimodal), local vs Hub paths, revisions, uploading to…
server-inference
Use this skill when the user wants to run or debug MLX-VLM server inference, including uv run mlxvlm.server, /v1/models, /v1/chat/completions, /v1/responses, streaming, OpenAI-compatible clients, health checks, metrics, model unload/reload, adapters, trust-remote-code, and server request/response failures.
cli-inference
Use this skill when the user wants to run or debug MLX-VLM inference from the command line, including uv run mlxvlm.generate, image/audio/video inputs, local model paths, Hugging Face model IDs, deterministic repro commands, and CLI errors around processors, prompts, model loading, or missing weights.
fda-database
Query openFDA API for drugs, devices, adverse events, recalls, regulatory submissions (510k, PMA), substance identification (UNII), for FDA regulatory data analysis and safety research.
meta-compliance-audit-bundle
Auditable compliance bundle: deep-research with citations → signable .docx report → read-only PDF archive → memory note of audit findings.