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 ask-openmedgit 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/ask-openmed)<a href="https://agentmods.dev/skills/maziyarpanahi/openmed/ask-openmed"><img src="https://agentmods.dev/badge/skills/maziyarpanahi/openmed/ask-openmed/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/ask-openmed"><img src="https://agentmods.dev/badge/skills/maziyarpanahi/openmed/ask-openmed.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 | $0.00069 | $0.02486 |
| Opus 5 | $0.00034 | $0.01243 |
| Sonnet 5 | $0.00014 | $0.00497 |
| Haiku 4.5 | $0.00007 | $0.00249 |
Grade A, and why
ask-openmed 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 10d 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 — 155 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ask OpenMed
Use this skill as the first pass when a request does not name a focused OpenMed workflow. It selects an existing skill identifier; it does not process input, download a model, call a service, or make a clinical decision. No mandatory network call is part of this route.
Deterministic routing contract
Apply these rules in order:
- Normalize only the stated goal: lowercase it, trim surrounding whitespace, and collapse repeated whitespace. Do not inspect or copy the data payload.
- Apply the intake boundary. If the goal matches an Intake row, select the first matching intake skill even when the goal also names a later stage. If sensitivity is unclear, its next handoff is the privacy gate: deidentifying-clinical-text, before extraction, exchange, or verification sees the parsed content.
- For a goal with no intake cue, apply the privacy override. If it asks to
analyze, extract, exchange, share, upload, or verify clinical or personal
content and is not explicitly marked safe, start at the privacy gate. The
only safe markers are
synthetic input,synthetic note,synthetic record,synthetic dataset,input is synthetic,data is synthetic,already de-identified, andalready deidentified. A marker does not count whenno,not,never,unknown, oruncertainoccurs in the four normalized words before it. Treat every ambiguous safety statement as sensitive. - Otherwise inspect the remaining tables in the fixed handoff order shown below. Treat each comma-separated cue as a case-insensitive substring of the normalized goal, and select the first matching row. Every cue cell is a comma-separated list of alternatives; conjunctions have no special meaning. Do not add synonyms or infer cues from the payload. This stage order and row order break every tie.
- If no row matches, use building-with-openmed as the orientation fallback.
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.
- 10d ago First seen · 155 lines · 69 tokens per session scan A 1ec9cc72ad18
ask-openmed is a skill published in the GitHub repository maziyarpanahi/openmed (5,282 stars, last pushed today), licensed Apache-2.0. It adds 69 tokens to every session and 2,486 once invoked, about $0.0003 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
add-new-model
Use this skill when the user wants to add or port a new model architecture to MLX-VLM — mapping a Hugging Face modeltype to a new file under mlxvlm/models, writing the ModelConfig, matching layer/weight names, reusing a similar existing model, adding a test class, and validating the port. Covers vision-language…
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…
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.
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.
contributing
Use this skill when the user wants to contribute to MLX-VLM — opening a PR, where model code/config/tests go, backward-compatible config args, running the test suite, code formatting and the pre-commit hooks (black, clang-format), and PR expectations (tests, review, perf evidence). Use it to set up a change so it…