clalit-mcp
CLI/MCP for your own Clalit (כללית) labs, prescriptions, and lab orders. Login on your machine, read-only.
Install / Use
claude mcp add netanelavr -- npx -y github:netanelavr/clalit-mcpIf the server publishes to npm under a different name, use that package instead — check the repo README.
MCP Server
Model Context Protocol server
Quality Score
Category
Development & EngineeringSupported Platforms
Tags
Our assessment of clalit-mcp
clalit-mcp scores 80/100 on our quality scale, 3565th of 4,702 Development & Engineering skills we index.
Its MCP Server is 7.6 KB long, well organised into 13 sections with 8 code examples: a thorough specification that gives an agent plenty to work with.
It has 10 GitHub stars, so there is little community track record yet; judge it on its content.
Maintenance, license and trust
- The repository was last updated yesterday, so clalit-mcp is actively maintained.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 97/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.
AI review by kimi-k2.7-code on 2026-11-17. Automated pattern scan on 2026-11-17. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.
clalit-mcp compared with similar skills
All 4 of these similar skills score higher than clalit-mcp; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| clalit-mcp (this skill)by netanelavr | 80 | 10 | 1d ago | MCP Server |
| taste-skillby Leonxlnx | 100 | 117.6k | 9d ago | CLAUDE.md |
| Agent-Reachby Panniantong | 100 | 112.7k | 9d ago | CLAUDE.md |
| rufloby ruvnet | 100 | 82.0k | today | CLAUDE.md |
| headroomby headroomlabs-ai | 100 | 79.9k | today | CLAUDE.md |
Frequently asked questions
- How do I install clalit-mcp?
- Run
claude mcp add netanelavr -- npx -y github:netanelavr/clalit-mcp. The install tabs above show the steps for each supported agent. - Which AI agents does clalit-mcp work with?
- It is written for Claude Code and Claude Desktop, as a MCP Server file. Other agents that read the same format can often use it too.
- Is clalit-mcp safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It is MIT-licensed and scores 97/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is clalit-mcp still maintained?
- The repository was last updated yesterday, so clalit-mcp is actively maintained.
Skill content
View source on GitHubClalit Health
Read your own Clalit records (lab results, prescriptions, lab orders) from a CLI or an AI assistant (MCP).
Unofficial. Not affiliated with Clalit. Not a substitute for the official portal or clinical advice.
The MCP server runs locally on your machine: you log in through your own browser, the session is stored only on your computer, and this project has no hosted backend and collects nothing. Note that whatever results your AI client (such as Claude) requests are passed to that client and are handled under its provider's privacy terms.

Getting started with an AI assistant
Copy and paste this to your agent:
Read my Clalit lab results from the last year and summarize anything out of range, then list my active prescriptions.
https://raw.githubusercontent.com/netanelavr/clalit-mcp/main/skills/clalit-mcp/SKILL.md
Or install the skill locally:
npx skills add netanelavr/clalit-mcp
Terms of use
- Own account only. Never use this against someone else's record.
- Read-only. No booking, payments, profile writes, or prescription requests.
- No CAPTCHA / bot / Imperva bypass. You solve CAPTCHA and enter the SMS OTP yourself.
- Residential machine. Datacenter and many cloud IPs get Imperva Error 16; log in on your own Mac / home network.
- Session file is secret. Stored mode
0600under~/.config/clalit-mcp/. Never commit or paste it.
Install
git clone https://github.com/netanelavr/clalit-mcp.git
cd clalit-mcp
npm install
npm test
Sign in and CLI
npx tsx packages/cli/src/main.ts login
npx tsx packages/cli/src/main.ts labs --json
npx tsx packages/cli/src/main.ts lab --ref <token-from-labs> --json
npx tsx packages/cli/src/main.ts lab-document --ref <token> --out result.pdf
npx tsx packages/cli/src/main.ts prescriptions --json
npx tsx packages/cli/src/main.ts lab-orders --json
npx tsx packages/cli/src/main.ts lab-order --ref <token-from-lab-orders> --json
npx tsx packages/cli/src/main.ts documents --source visit_summaries --json # or hospital_summaries | medical_approvals | form17
npx tsx packages/cli/src/main.ts document --ref <token-from-documents> --out summary.pdf
npx tsx packages/cli/src/main.ts records --source diagnoses --json # sensitivities | chronic_medications | medicines_purchased | medication_approvals | imaging_clinics | imaging_hospitals | imaging_external | hospital_visits | hospital_labs | referrals | appointments | recommendations | vaccinations | blood_type
npx tsx packages/cli/src/main.ts hospital-lab-result --link <link-from-hospital_labs-row> --hospital <its-context.value> --json
npx tsx packages/cli/src/main.ts record-document --ref <token-from-records> --out report.pdf
npx tsx packages/cli/src/main.ts export --out ~/health
npx tsx packages/cli/src/main.ts help
login opens a loopback browser page for ID, CAPTCHA, and SMS OTP. See docs/LIVE-VERIFY.md. login --browser opens your own Chrome at Clalit Online's sign-in page instead: sign in there as you usually do — with your permanent password if you have one, so no SMS — and the signed-in cookies become the session (docs/AUTH.md).
Keep a copy of everything: export
npx tsx packages/cli/src/main.ts export # ./clalit-export-<today>/: labs (every result and its PDF), prescriptions, lab orders, documents, records
npx tsx packages/cli/src/main.ts export --out ~/health # under that directory; --no-documents skips the PDFs
One folder: labs (one JSON per result, plus its PDF), prescriptions, lab orders, documents (visit summaries, hospital
summaries, medical approvals and Form 17 — a list and the PDFs per source), records (diagnoses, sensitivities, the
regular-medication list, pharmacy purchases, medication approvals, imaging reports with their PDFs, hospital visits and
labs with every draw's results, referrals with their PDFs, upcoming appointments, the portal's own recommendations,
vaccinations, blood type), and a root README.md that describes each folder for an AI assistant (layout in docs/CLI.md). --from/--to (dd.MM.yyyy) filter labs only, default 01.01.1990 to today; a rerun skips what it already has.
Layout modeled on Health Records Export for Maccabi; liberate-labs can load it into SQLite.
The folder is not encrypted, and anything you open in an AI app is sent to that provider (see the privacy note at the top).
MCP
Sign in first (login above); the MCP server reuses that local session.
Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS). Use absolute paths (Dock-launched Claude does not see nvm / shell PATH):
{
"mcpServers": {
"clalit": {
"command": "/absolute/path/to/node",
"args": [
"/absolute/path/to/clalit-mcp/node_modules/.bin/tsx",
"/absolute/path/to/clalit-mcp/packages/cli/src/main.ts",
"mcp"
]
}
}
}
Find your node path with which node. Quit Claude fully (Cmd+Q) and reopen.

Other MCP clients (Cursor, Claude Code, …)
{
"mcpServers": {
"clalit": {
"command": "npx",
"args": ["tsx", "packages/cli/src/main.ts", "mcp"],
"cwd": "/absolute/path/to/clalit-mcp"
}
}
}
Signing in from the assistant
A session made with login is picked up. Without one, the tools answer with a link to a sign-in page the server serves itself on http://localhost:8765/ (clalit_login returns the same link): open it, pick how long to stay signed in (1 hour to 1 week), and enter the ID, the CAPTCHA and the SMS code there — only ever on that page, never in the chat. While the server runs it keeps the session alive and signs out at the end of the length you picked (docs/MCP.md).
Tools
| Tool | What it reads |
| --- | --- |
| list_labs, get_lab_result, get_lab_document | Lab history, one result, the original lab PDF |
| list_prescriptions, get_prescription_issue_status | Prescriptions and their pharmacy issue status |
| list_lab_orders, get_lab_order | Lab orders and one order's detail |
| list_documents, get_document | Visit summaries, hospital summaries, medical approvals, Form 17, and their PDFs |
| list_records, get_record_document, get_hospital_lab_result | Diagnoses, sensitivities, chronic medications, purchased medicines, medication approvals, imaging reports (clinics, hospitals, external) with their PDFs, hospital visits, hospital labs and their result pages, referrals with their PDFs, upcoming appointments, the portal's recommendations, vaccinations, blood type |
| clalit_login, clalit_session_status, clalit_logout | The sign-in link, whether a session is live and until when, and sign-out |
Mapping the portal: tools/sitemap.mts
npx tsx tools/sitemap.mts --out ./clalit-sitemap # sitemap.json: which pages your account has and what they are made of
npx tsx tools/sitemap.mts --out ./clalit-sitemap --raw # also keeps each page's HTML under raw/ (private) for writing parsers
A GET-only, paced walk over the signed-in portal's own links. sitemap.json records titles, headings, form field names,
dropdown labels and grid headers — never a value — so a page nobody reads yet can be turned into a reader from evidence.
The record sources above were built from one such map.
License
Related Skills
taste-skill
117.6kAgent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
Agent-Reach
112.7kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
ruflo
82.0k🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
headroom
79.9kCompress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
