Chia Health MCP Server
health.chia/chia-mcpchia.healthstreamable-httpv1.0.0Scanned
Licensed US telehealth — GLP-1 medications, intake, consents, Stripe ACP. HIPAA-compliant, 30 tools.
34 tools: 17 can delete or cancel things, 1 do something that can't be undone, 3 change things and 13 only read.
- Reads private data · 4
- Reads text others write · 8
- Can send data out · 2
- 5 findings
Tools listed without sign-in. Remote posture
With Anlyon
17 tools here can delete or cancel things and 1 does something that can't be undone. Govern them.
- Anlyon lists the tools and adopts none.
- Each tool you adopt gets approvals and receipts.
- Your agent never holds the key.
A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.
What the Chia Health MCP Server can do
The Chia Health MCP Server is published by chia.health, a domain the MCP Registry verified. It has 34 tools. It can delete or cancel things (17 tools: auth_start, auth_verify_otp, auth_resend_otp, intake_submit and 13 more), do things that cannot be undone, such as sending, paying or publishing (auth_check_payment) and create or change data (3 tools: medications_availability, eligibility_check and checkout_status). 13 tools only read.
4 tools read private data (consent_list, consent_text, order_documents and 1 more); 8 tools bring in text other people can write (medications_list, medications_details, medications_availability and 5 more); 2 tools can send data out of your system (auth_resend_otp and portal_message). On its own it has all three parts of a prompt-injection exfiltration path: private data, text others write, and a way out.
The fixed checks matched 5 findings (annotation-conflict and long-description). Each one comes with the text it matched.
Its remote endpoint lists its tools without sign-in, and its tools changed once since Anlyon Radar first saw them.
Last scanned Nov 11, 2026, version 1.0.0.
Tools
Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.
34 tools
Deletes or cancels 17Claude Code: deny
auth_resend_otpResend the verification code to the patient's email.Can send data outNo findings
Why it is in this group
- The name "auth_resend_otp" says "resend": a word for sending, and what is sent cannot be taken back.
- The server marks it destructive (destructiveHint: true).
What it can reach
- outboundThe name "auth_resend_otp" says "resend": it sends or publishes something to other people.
Description, as the server wrote it Resend the verification code to the patient's email. Use this if the original code expired (5-minute window) or was not received. Requires the session_id from auth_start — no email needed.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: false
auth_startStart a patient session by providing their contact information.No findings
Why it is in this group
- The name "auth_start" says "start": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Start a patient session by providing their contact information. Sends a 6-digit verification code to the patient's email. Returns a session_id (NOT a token). The session_id is used with auth_verify_otp to prove email ownership and get a bearer token. The code is in the email subject line: 'Chia Health: Your code is XXXXXX'. If you have access to the patient's email (e.g. Gmail MCP), search for this subject. No authentication required. Call this when the patient is ready to proceed with their medical intake — after browsing medications and checking eligibility.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: false
auth_verify_otpVerify the 6-digit code sent to the patient's email.No findings
Why it is in this group
- The input takes a statement or command to run (input.code), which can change anything it can reach.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Verify the 6-digit code sent to the patient's email. Returns a guest-scope bearer token for intake, consent, order, and checkout tools. Requires the session_id from auth_start — no email needed. After checkout and payment, call auth_check_payment to upgrade the token to full scope for portal access.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: false
checkout_cancelCancel an in-progress checkout session.No findings
Why it is in this group
- The name "checkout_cancel" says "checkout": a word for a change.
- The description's first verb is "cancel": a word for removing or ending something.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Cancel an in-progress checkout session. Releases any held inventory and cancels the associated Stripe PaymentIntent if not yet captured. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: true
checkout_completeComplete payment using Stripe ACP (Shared Payment Token).No findings
Why it is in this group
- The name "checkout_complete" says "checkout": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Complete payment using Stripe ACP (Shared Payment Token). Only use this if your platform supports Stripe Agentic Commerce Protocol and can provision an SPT. If your platform does NOT support ACP, use the `payment_url` from checkout_create instead, then poll checkout_status. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: true
checkout_createInitiate a checkout session for a medication order.No findings
Why it is in this group
- The name "checkout_create" says "checkout": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Initiate a checkout session for a medication order. Returns checkout details including line items, total, and payment options. TWO PAYMENT PATHS are supported: 1. **Stripe ACP (preferred)**: If your platform supports Stripe Agentic Commerce Protocol, provision a Shared Payment Token (SPT) and call checkout_complete to pay instantly. 2. **Payment link (fallback)**: If ACP/SPT is not available, present the returned `payment_url` to the patient. This is a Stripe-hosted checkout page where the patient can enter their card and pay directly. After sending the link, call checkout_status to poll for payment completion. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: true
checkout_updateUpdate an existing checkout session.No findings
Why it is in this group
- The name "checkout_update" says "checkout": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Update an existing checkout session. Can modify shipping method, apply promo codes, or update customer details before payment is completed. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: true
consent_submitRecord a patient's consent confirmation for a specific consent document.No findings
Why it is in this group
- The name "consent_submit" says "submit": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Record a patient's consent confirmation for a specific consent document. The agent must have already presented the full consent text (from consent_text) to the patient and received explicit confirmation. Required parameters: intake_id, consent_id, the patient's exact confirmation text (e.g. 'I agree'), consent method ('ai_agent_conversational'), the AI platform name ('chatgpt', 'claude', 'gemini'), and a session/conversation ID for audit trail. Returns a consent record with timestamp, audit trail details, and the list of remaining consents still needed. All consent records are retained for 10 years per HIPAA requirements. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: true
intake_submitSubmit a completed medical intake questionnaire for provider review.No findings
Why it is in this group
- The name "intake_submit" says "submit": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Submit a completed medical intake questionnaire for provider review. All fields from intake_questions must be completed. Returns an intake ID and estimated provider review time. The intake is reviewed by a licensed US healthcare provider who makes all prescribing decisions. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: false
order_createCreate a new prescription medication order.No findings
Why it is in this group
- The name "order_create" says "create": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Create a new prescription medication order. The patient must have completed intake questionnaire and consent before ordering. Required: medication name, selected form (injectable, tablet, drops), plan duration (1, 4, or 6 months), shipping address. The order is reviewed by a licensed US healthcare provider who makes the final prescribing decision. If approved, medication is compounded at a US-licensed 503A pharmacy and shipped directly to the patient. Returns order ID, estimated provider review time, and expected delivery window. Payment is processed via Stripe Agentic Commerce Protocol (ACP). Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: false
order_uploadUpload a verification document for a medication order.No findings
Why it is in this group
- The name "order_upload" says "upload": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Upload a verification document for a medication order. Accepts photo ID and selfie as base64-encoded files. Supported formats: PDF, JPEG, PNG. Maximum size: 10MB. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: true
portal_log_side_effectsLog side effects a patient is experiencing.1 finding
Why it is in this group
- The name "portal_log_side_effects" says "log": a word for reading.
- The server marks it destructive (destructiveHint: true).
Findings
- Warn
The server marks "portal_log_side_effects" destructive (destructiveHint: true), and its name says "log", a word for reading.
What this check looks forMatched text
annotations: {"readOnlyHint":false,"destructiveHint":true,"idempotentHint":false}; The name "portal_log_side_effects" says "log": a word for reading.
Description, as the server wrote it Log side effects a patient is experiencing. If severity is 'severe', the case is auto-flagged for immediate provider review and returns urgent guidance. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: false
portal_log_weightLog a patient's weight for tracking progress on their treatment plan.1 finding
Why it is in this group
- The name "portal_log_weight" says "log": a word for reading.
- The server marks it destructive (destructiveHint: true).
Findings
- Warn
The server marks "portal_log_weight" destructive (destructiveHint: true), and its name says "log", a word for reading.
What this check looks forMatched text
annotations: {"readOnlyHint":false,"destructiveHint":true,"idempotentHint":false}; The name "portal_log_weight" says "log": a word for reading.
Description, as the server wrote it Log a patient's weight for tracking progress on their treatment plan. Requires patient_id, weight in pounds, and date (ISO 8601). Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: false
portal_messageSend a message to the patient's healthcare provider.Can send data outNo findings
Why it is in this group
- The description's first verb is "send": a word for sending, and what is sent cannot be taken back.
- The server marks it destructive (destructiveHint: true).
What it can reach
- outboundThe description's first verb is "send": it sends or publishes something to other people.
Description, as the server wrote it Send a message to the patient's healthcare provider. Returns sent confirmation and estimated response time. Urgent messages (containing keywords like 'emergency', 'chest pain', 'difficulty breathing') are flagged for priority response. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: false
portal_refillRequest a medication refill for the patient's current prescription.No findings
Why it is in this group
- The description's first verb is "request": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Request a medication refill for the patient's current prescription. Creates a refill order that will be reviewed by a provider within 24-48 hours. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: false
portal_supportContact customer support with a question or issue.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Contact customer support with a question or issue. Creates a support ticket and returns the ticket ID and estimated response time. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: false
provider_respondSubmit answers to provider follow-up questions for a specific order.No findings
Why it is in this group
- The description's first verb is "submit": a word for a change.
- The server marks it destructive (destructiveHint: true).
Description, as the server wrote it Submit answers to provider follow-up questions for a specific order. The responses are sent to the provider for review. Returns confirmation and updated order status. Requires authentication.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: true
- idempotentHint: true
Can't be undone 1Claude Code: ask
auth_check_paymentCheck if the patient has completed payment and upgrade the token to full scope.1 finding
Why it is in this group
- The name "auth_check_payment" says "payment": a word that usually means money moves, which cannot be taken back.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
The server marks "auth_check_payment" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
What this check looks forMatched text
annotations: {"readOnlyHint":true}; The name "auth_check_payment" says "payment": a word that usually means money moves, which cannot be taken back.
Description, as the server wrote it Check if the patient has completed payment and upgrade the token to full scope. Call this after sharing the checkout payment link with the patient. Poll every 10-15 seconds. When payment is detected, the token is automatically upgraded to full scope, unlocking portal tools (care plan, refills, messaging, weight logs, etc.). Requires the guest token from auth_verify_otp as bearer_token.
Annotations the server declares
- readOnlyHint: true
Changes things 3Claude Code: ask
checkout_statusCheck the payment status of a checkout session.1 finding
Why it is in this group
- The name "checkout_status" says "checkout": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
The server marks "checkout_status" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
What this check looks forMatched text
annotations: {"readOnlyHint":true}; The name "checkout_status" says "checkout": a word for a change.
Description, as the server wrote it Check the payment status of a checkout session. Use this to poll for completion after sending the patient a payment link (the `payment_url` from checkout_create). When the patient pays via the link, this tool detects the payment, triggers order fulfillment, and returns the confirmation. Poll every 5-10 seconds while waiting. Requires authentication.
Annotations the server declares
- readOnlyHint: true
eligibility_checkPre-screen a patient's basic eligibility for telehealth prescription services.Reads text others writeNo findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it Pre-screen a patient's basic eligibility for telehealth prescription services. Required: age (18+) and state (where the patient resides). Optional: BMI (20+ required for GLP-1 / weight-loss products), biological sex, pregnancy status, and diagnosed conditions. Only pass parameters that apply to this patient. `pregnancy_status` applies ONLY when biological sex is female — omit it entirely for males. Don't invent values to satisfy the schema; if you don't know, leave the parameter out and the server will return what is or isn't checkable. If you already know the patient's age, sex, state, height/weight from prior conversation context, you may pre-fill — but read the values back to the patient and get explicit confirmation before calling this tool. Returns eligibility status, available medications, and any disqualifying reasons (MTC/MEN2 history, pregnancy, out-of-coverage state, etc.).
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
medications_availabilityCheck if a specific medication is available for shipping to the patient's state.Reads text others writeNo findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it Check if a specific medication is available for shipping to the patient's state. Some compounded medications have state-specific restrictions based on pharmacy licensing. Returns availability status and reason if unavailable; when the name fits more than one treatment (e.g. a medication sold in two forms), 'options' gives each one's status.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
Only reads 13Show 13 read-only tools
consent_listGet the list of all consent documents a patient must accept before ordering medication.Reads private dataNo findings
Why it is in this group
- The name "consent_list" says "list": a word for reading.
What it can reach
- private-dataThe description's first sentence says "get" and "document": it reads documents.
Description, as the server wrote it Get the list of all consent documents a patient must accept before ordering medication. Returns consent IDs, titles, summaries, and order of presentation. Required consents include: telehealth informed consent, compounded medication treatment consent, pharmacy authorization, HIPAA notice of privacy practices, and AI-assisted intake disclosure. Each consent must be fetched individually via consent_text and confirmed by the patient before proceeding. Requires authentication.
Annotations the server declares
- readOnlyHint: true
consent_statusCheck whether all required consents are complete for a patient intake.No findings
Why it is in this group
- The name "consent_status" says "status": a word for reading.
Description, as the server wrote it Check whether all required consents are complete for a patient intake. Returns status of each consent and whether the patient can proceed to ordering. This is a gate — order_create will reject if consents are incomplete. Requires authentication.
Annotations the server declares
- readOnlyHint: true
consent_textFetch the full text of a specific consent document for patient review.Reads private dataReads text others writeNo findings
Why it is in this group
- The description's first verb is "fetch": a word for reading.
What it can reach
- private-dataThe description's first sentence says "fetch" and "document": it reads documents.
- untrusted-inputThe description's first sentence says "fetch" and "review": it reads comments, which other people can write.
Description, as the server wrote it Fetch the full text of a specific consent document for patient review. Returns the complete consent document split into titled sections that the agent MUST present to the patient verbatim in the conversation — do not summarize or paraphrase. Includes: consent version number, effective date, section headings and body text, a confirmation prompt the patient should agree to, and withdrawal instructions. Available consent types: telehealth informed consent, compounded medication treatment consent, pharmacy authorization, HIPAA notice of privacy practices, and AI-assisted intake disclosure. The patient must explicitly confirm each consent before the agent can call consent_submit. Requires authentication.
Annotations the server declares
- readOnlyHint: true
intake_questionsGet the medical intake questionnaire for the chosen medication(s).Reads text others write1 finding
Why it is in this group
- The description's first verb is "get": a word for reading.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Findings
- Info
The description is 2,041 characters long, over the 2,000 this check looks for.
What this check looks forMatched text
description ends: "...d US healthcare provider reviews both phases and makes all prescribing decisions."
Description, as the server wrote it Get the medical intake questionnaire for the chosen medication(s). The questionnaire is product-aware: GLP-1 / weight-loss medications return weight-loss goals, GLP-1 history, and MTC/MEN2 screening; NAD+ and other longevity peptides return energy/sleep/stress/cognitive/delivery-method questions instead. If the patient wants more than one medication, pass the additional slugs in `additional_medications` — the server returns the UNION of section sets deduped by section key, so you ask each shared question exactly once. ## How to present this to the patient 1. PROGRESSIVE DISCLOSURE: walk through ONE section at a time. Wait for the patient's reply before moving to the next section. Do not paste the whole questionnaire in a single message. 2. HONOR CONDITIONALS: each section and each question may carry a `conditional_on` predicate (e.g. `{sex_assigned_at_birth: Female}` on the Pregnancy section). SKIP any section/question whose predicate isn't satisfied. Don't ask males about pregnancy or perimenopause. 3. QUIZ FORMAT: present every `select` / `multi_select` question as a short pick-list using the `options` array verbatim. The patient should be able to reply with a single choice, not a sentence. Reserve free text for `*_details` follow-ups. 4. EASY FIRST: order sections from low-friction (goals, lifestyle, preferences) to high-friction (clinical history, MTC/MEN2, prior therapies). The provider sees all answers regardless of order asked. 5. USE-AND-VERIFY: if you know answers from prior conversation context, pre-fill them in your draft, but read them back to the patient and get explicit OK before calling `intake_submit`. Never silently submit assumed values. Returns two phases: (1) pre_checkout — eligibility / screening questions, collected and submitted BEFORE payment; (2) post_checkout — detailed clinical history, collected and submitted AFTER payment. Do not submit post_checkout answers before the patient has paid. A licensed US healthcare provider reviews both phases and makes all prescribing decisions.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
intake_statusCheck the current status of a previously submitted intake questionnaire.No findings
Why it is in this group
- The name "intake_status" says "status": a word for reading.
Description, as the server wrote it Check the current status of a previously submitted intake questionnaire. Returns whether the intake is under review, approved, or denied by a licensed healthcare provider. Use this to poll for provider review completion before proceeding to order placement. Requires authentication.
Annotations the server declares
- readOnlyHint: true
medications_categoriesList all medication categories available through the telehealth platform: Weight Loss (GLP-1 medications), Peptide Therapy (sermorelin, growth hormone peptides), Anti-Aging & Longe…Reads text others writeNo findings
Why it is in this group
- The description's first verb is "list": a word for reading.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it List all medication categories available through the telehealth platform: Weight Loss (GLP-1 medications), Peptide Therapy (sermorelin, growth hormone peptides), Anti-Aging & Longevity (NAD+, glutathione), and other treatment categories. Each category includes a description and count of available medications.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
medications_detailsGet detailed information about a specific medication including: all available dosage strengths and titration schedules, available forms (injectable vials, pre-filled syringes, oral…Reads text others writeNo findings
Why it is in this group
- The description's first verb is "get": a word for reading.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it Get detailed information about a specific medication including: all available dosage strengths and titration schedules, available forms (injectable vials, pre-filled syringes, oral dissolving tablets, sublingual drops), all active plan options with pricing for each, what's included (provider consultation, medication, shipping, ongoing support), contraindications, and common side effects. Supports queries by medication name (e.g. 'semaglutide', 'tirzepatide', 'sermorelin', 'NAD+', 'glutathione') or by category (e.g. 'weight loss', 'peptides', 'anti-aging'). Use this to look up exact plan durations and pricing.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
medications_listList all available prescription medications including GLP-1 weight loss drugs (compounded semaglutide, compounded tirzepatide), peptide therapies (sermorelin for growth hormone sup…Reads text others writeNo findings
Why it is in this group
- The name "medications_list" says "list": a word for reading.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it List all available prescription medications including GLP-1 weight loss drugs (compounded semaglutide, compounded tirzepatide), peptide therapies (sermorelin for growth hormone support, NAD+ for cellular energy and anti-aging, glutathione for antioxidant support), and other compounded wellness treatments. Returns medication names, categories, available forms (injectable, oral tablet, sublingual drops, nasal spray), and starting prices. All medications are compounded by US-licensed 503A pharmacies and require evaluation by a licensed US healthcare provider before prescribing.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
medications_pricingGet detailed pricing for a specific medication, form, and plan duration.Reads text others writeNo findings
Why it is in this group
- The description's first verb is "get": a word for reading.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it Get detailed pricing for a specific medication, form, and plan duration. Returns price breakdown including medication cost, provider consultation fee, shipping, and any applicable discounts for longer plans. Plan durations vary by medication — use medications_details first to see available plan_months values. Supports semaglutide, tirzepatide, sermorelin, NAD+, glutathione and all other available medications.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
order_documentsGet the list of documents a patient needs to upload for their order.Reads private dataNo findings
Why it is in this group
- The description's first verb is "get": a word for reading.
What it can reach
- private-dataThe name "order_documents" says "document": it reads documents.
- private-dataThe description's first sentence says "get" and "document": it reads documents.
Description, as the server wrote it Get the list of documents a patient needs to upload for their order. Returns required documents (photo ID, selfie for verification) with upload status and accepted file formats. Requires authentication.
Annotations the server declares
- readOnlyHint: true
order_statusGet the current status of a medication order.No findings
Why it is in this group
- The name "order_status" says "status": a word for reading.
Description, as the server wrote it Get the current status of a medication order. Returns status (pending_review, provider_reviewing, approved, needs_info, denied, compounding, shipped, delivered), tracking information, and delivery estimate. Requires authentication.
Annotations the server declares
- readOnlyHint: true
portal_care_planGet the patient's current care plan including: current medication, current dosage, titration schedule, next dose adjustment date, upcoming refill date, provider notes, and weight p…Reads private dataNo findings
Why it is in this group
- The description's first verb is "get": a word for reading.
What it can reach
- private-dataThe description's first sentence says "get" and "note": it reads notes.
Description, as the server wrote it Get the patient's current care plan including: current medication, current dosage, titration schedule, next dose adjustment date, upcoming refill date, provider notes, and weight progress summary. Requires authentication.
Annotations the server declares
- readOnlyHint: true
provider_questionsGet follow-up questions from the healthcare provider for a specific order.No findings
Why it is in this group
- The description's first verb is "get": a word for reading.
Description, as the server wrote it Get follow-up questions from the healthcare provider for a specific order. The provider may request additional information before making a prescribing decision. Returns the questions if the order status is 'needs_info', or a message that no questions are pending. Requires authentication.
Annotations the server declares
- readOnlyHint: true
This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.
Exfil surface
Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.
64 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 2 outbound tools breaks every one.
- Reads private data4
- Reads files, mail, messages, databases, secrets, calendars or repositories.
- Reads text others write8
- Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
- Can send data out2
- Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
{
"permissions": {
"deny": [
"mcp__chia__auth_resend_otp",
"mcp__chia__portal_message"
]
}
}Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.
Remote posture
What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.
Looked at Nov 20, 2026, 18:54 UTC, unchanged since Oct 22, 2026. No credentials were sent and no tool was called. What is looked at.
- Reachable
- yes
- HTTP status of an anonymous initialize
- 200
- tools/list served without credentials
- yes
- Tools that change things, listed without credentials
- 21
- Protocol version
- 2025-06-18
- WWW-Authenticate on 401
- not observed
- resource_metadata in WWW-Authenticate
- not observed
- RFC 9728 protected resource metadata
- not observed
- Authorization servers
- not observed
- Legacy HTTP+SSE transport
- no
- Dynamic client registration
- not observed
- Client ID metadata documents
- not observed
- HighThe whole server
The endpoint served tools/list without credentials, and the list has 21 tools that change things (17 destructive, 1 irreversible, 3 write). Whether a call also works without credentials was not tested: tools are never called.
What this check looks forMatched text
https://mcp.chia.health/: tools/list returned 34 tools to an anonymous client, including auth_start (destructive), auth_verify_otp (destructive), auth_resend_otp (destructive), intake_submit (destructive), consent_submit (destructive) and 16 more
Findings
Matches of fixed checks, with the text they matched.
The server marks "auth_check_payment" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
What this check looks forMatched text
annotations: {"readOnlyHint":true}; The name "auth_check_payment" says "payment": a word that usually means money moves, which cannot be taken back.
- Warncheckout_status
The server marks "checkout_status" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
What this check looks forMatched text
annotations: {"readOnlyHint":true}; The name "checkout_status" says "checkout": a word for a change.
The server marks "portal_log_weight" destructive (destructiveHint: true), and its name says "log", a word for reading.
What this check looks forMatched text
annotations: {"readOnlyHint":false,"destructiveHint":true,"idempotentHint":false}; The name "portal_log_weight" says "log": a word for reading.
The server marks "portal_log_side_effects" destructive (destructiveHint: true), and its name says "log", a word for reading.
What this check looks forMatched text
annotations: {"readOnlyHint":false,"destructiveHint":true,"idempotentHint":false}; The name "portal_log_side_effects" says "log": a word for reading.
- Infointake_questions
The description is 2,041 characters long, over the 2,000 this check looks for.
What this check looks forMatched text
description ends: "...d US healthcare provider reviews both phases and makes all prescribing decisions."
History
Changes to its tool definitions and to its remote endpoint, newest first.
Tools3 changed.~3 changedView diff
- Changed:medications_availabilityChanges thingsdescriptioninputSchema
- Changed:medications_pricingOnly readsinputSchema
- Changed:order_createDeletes or cancelsinputSchema
Install
Pinned to version 1.0.0. Values in angle brackets are yours to fill in.
claude mcp add --transport http chia https://mcp.chia.health/claude mcp add --transport http chia https://mcp.chia.health/Run in a terminal.
Claude Code permissions
Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.
{
"permissions": {
"allow": [
"mcp__chia__medications_list",
"mcp__chia__medications_details",
"mcp__chia__medications_pricing",
"mcp__chia__medications_categories",
"mcp__chia__intake_questions",
"mcp__chia__intake_status",
"mcp__chia__consent_list",
"mcp__chia__consent_text",
"mcp__chia__consent_status",
"mcp__chia__order_status",
"mcp__chia__order_documents",
"mcp__chia__portal_care_plan",
"mcp__chia__provider_questions"
],
"ask": [
"mcp__chia__auth_check_payment",
"mcp__chia__medications_availability",
"mcp__chia__eligibility_check",
"mcp__chia__checkout_status"
],
"deny": [
"mcp__chia__auth_start",
"mcp__chia__auth_verify_otp",
"mcp__chia__auth_resend_otp",
"mcp__chia__intake_submit",
"mcp__chia__consent_submit",
"mcp__chia__order_create",
"mcp__chia__order_upload",
"mcp__chia__checkout_create",
"mcp__chia__checkout_update",
"mcp__chia__checkout_complete",
"mcp__chia__checkout_cancel",
"mcp__chia__portal_log_weight",
"mcp__chia__portal_log_side_effects",
"mcp__chia__portal_message",
"mcp__chia__portal_refill",
"mcp__chia__portal_support",
"mcp__chia__provider_respond"
]
}
}Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as chia. Review it before you use it.
Capability tags behind 13 rules
- mcp__chia__auth_resend_otp outbound
- mcp__chia__medications_list untrusted-input
- mcp__chia__medications_details untrusted-input
- mcp__chia__medications_availability untrusted-input
- mcp__chia__medications_pricing untrusted-input
- mcp__chia__medications_categories untrusted-input
- mcp__chia__eligibility_check untrusted-input
- mcp__chia__intake_questions untrusted-input
- mcp__chia__consent_list private-data
- mcp__chia__consent_text private-data, untrusted-input
- mcp__chia__order_documents private-data
- mcp__chia__portal_message outbound
- mcp__chia__portal_care_plan private-data
Pin
Keep the decision: fail CI when the tools change, follow every change, or show a badge.
Pin in CI
Fail a build when the tools you reviewed are not the tools the server serves today.
curl -s "https://radar.anlyon.com/api/v1/servers/health.chia/chia-mcp/verify?hash=2e1511eb9afdd1442a7e3bc1196a93f8aadc4645bb42b26bdc2297948f3293c1" | jq -e '.changed == false'Pinned hash 2e1511eb9afd, from Nov 11, 2026, 14:18 UTC.
Watch
Get told when its tools change: a new tool, a new description, a new class.
https://radar.anlyon.com/servers/health.chia/chia-mcp/changes.xmlAn Atom feed. Works in any feed reader or Slack's RSS app.
Badge
[](https://radar.anlyon.com/servers/health.chia/chia-mcp)Source: MCP Registry · repository. Synced Oct 21, 2026, 18:13 UTC.
Chia Health MCP Server: questions and answers
Answered from the scan above, not from the publisher's description.
Is the Chia Health MCP Server safe?
- No scan can promise that. What its definitions show: 34 tools, 21 of which can change, send or delete something, and 5 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.
Can the Chia Health MCP Server delete data?
- Yes. 17 tools can delete, cancel or revoke something: auth_start, auth_verify_otp, auth_resend_otp, intake_submit, consent_submit, order_create and 11 more.
Can the Chia Health MCP Server do things that cannot be undone?
- Yes. 1 tool can send, pay, publish or merge: auth_check_payment.
What tools does the Chia Health MCP Server have?
- 34 tools: 17 destructive, 1 irreversible, 3 write, 13 read. auth_start, auth_verify_otp, auth_resend_otp, intake_submit, consent_submit, order_create, order_upload, checkout_create and 26 more.
Does the Chia Health MCP Server need sign-in or OAuth?
- Its remote endpoint listed its tools to an anonymous client, without sign-in.
Does the Chia Health MCP Server send data to outside services?
- 2 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): auth_resend_otp and portal_message.
Has the Chia Health MCP Server changed its tools?
- Yes, once since Anlyon Radar started tracking it, most recently on Nov 11, 2026. Every change is on this page and in its Atom feed.