GymCoach
근력 트레이닝 기록 및 개인 AI 코치
아직 감지된 업데이트가 없습니다. GitHub 릴리스가 나오면 여기에 보입니다.
개인 계정의 공개 저장소 소유자는 GitHub로 로그인해 운영자를 확인하고 소개를 직접 관리할 수 있습니다. 관리자 승인은 필요하지 않습니다.
로그인한 뒤 확인 파일을 만들어 이 저장소의 기본 브랜치에 30분 안에 커밋해 주세요. 다른 브랜치나 포크의 파일은 인정하지 않습니다. 파일을 올린 뒤 아래에서 확인하면 관리자 승인 없이 관리할 수 있습니다. 자격을 유지하려면 파일을 남겨 두세요.
로그인이 필요하면 로그인 화면으로 이동합니다. 로그인 후 이곳으로 돌아와 원하는 확인 버튼을 다시 눌러 주세요.
프로젝트 폴더에서 /nomorevibe verify를 실행하면 소유권을 확인하고 소개를 직접 관리할 수 있습니다.
1분 만에 확인하기메이커가 쓴 소개는 위의 한 줄이 전부입니다. 더 자세한 내용은 저장소 README에서 읽을 수 있습니다. 저장소 README ↗
README 전체를 불러오는 중입니다.
README 전문 보기 ↗Drug evidence for busy physicians: search a drug, condition or mechanism and get, in three clicks or fewer, the published evidence, its GRADE certainty, who funded it, and whether the use is off-label. Every claim links to its source.
MARKETPLACE_API_KEY is set; get a free key at developer.cms.gov/marketplace-api); pharmacy cost per product (NADAC) with a 22-month price trend; Medicare Part D and Part B spending; Part B payment limits (ASP + 6%); VA Federal Supply Schedule prices; cheaper drugs in the same class; ways to save. Copy cost summary.| Workflow | When | What |
|---|---|---|
| Import cost data | Mondays 06:17 UTC, or Actions → Import cost data → Run workflow | Downloads the newest NADAC (and last year's, for the price trend), Medicare Part D and Part B Spending, the CMS monthly Part D formulary files (summarized by drug, insurer and state), Part B ASP payment limits, CMS negotiated prices and the VA price list into data/costs/ and commits them. Also records the Part D Prescribers API used by Specialists. If one source fails, last week's data is kept and the run is marked failed. |
| Live smoke test | Manual only | Builds the site and loads every tab against the real APIs, printing what each tab shows and any unavailable source. |
| Judgement | Rule | Code |
|---|---|---|
| Evidence certainty | GRADE starting certainty from study design. Randomized → High (one lone trial → Moderate for imprecision), observational → Low, case reports only → Very low. Preprints are shown but not counted. Risk of bias, inconsistency, indirectness and publication bias are listed as "not assessed". | src/lib/grade.ts |
| Off-label | The condition is matched against the FDA label's Indications and Usage text. Text under "Limitations of Use" never counts as approval. Synonyms can be missed, so the page always links the label. | src/lib/label-status.ts |
| Funding | 1) sponsor class of a linked ClinicalTrials.gov trial; 2) funding agencies on the PubMed record. Anything else is "Funding not reported", never a guess. | src/lib/funding.ts |
| "Credentialed journals" | PubMed medline[sb] (MEDLINE-indexed). Preprints come from Europe PMC (SRC:PPR) and are badged. | src/lib/sources/ |
npm install
cp .env.example .env.local # all keys optional
npm run dev # http://localhost:3000
npm run dev:demo # offline, with clearly marked placeholder data
npm test && npm run lint && npm run typecheck
NCBI_API_KEY and OPENFDA_API_KEY (both free), and later ANTHROPIC_API_KEY."use cache" with cacheLife("days")), so repeat searches are fast and stay within the free API limits.openFDA / DailyMed (labels, Drugs@FDA), PubMed E-utilities (MEDLINE literature), Europe PMC (preprints), ClinicalTrials.gov API v2 (trials, sponsors), Open Targets Platform (mechanisms, condition and target → drug), MedlinePlus Connect (patient information), Medicaid NADAC, CMS Medicare Part D and Part B Spending, CMS Part D formulary files, CMS ASP pricing files, CMS negotiated prices, the VA Federal Supply Schedule and the HealthCare.gov Marketplace API (costs and coverage), CMS Part D Prescribers by Provider and Drug and the NPPES NPI Registry (specialists).
No database and no accounts. Patient details in the letter tool live only in browser memory. Only public literature is sent to Claude.
Physician sign-in with NPI verification, which also enables self-reported experience on the Specialists tab; Medicare plan coverage from CMS formulary files; licensed retail price data for out-of-pocket estimates.