Connect
Optimize
Secure
The #1 agentic semantic tool search: 91.6% first-try accuracy on S1 Search Bench • Explore Tool Discovery →
Connect your AI agent to StackOne's Candid Health MCP server and give it 45 MCP tools out of the box. Auth, tool execution, and security all managed.
Coverage
Create, read, update, and delete across Candid Health — and extend your agent's capabilities with custom actions.
Authentication
Per-user OAuth in one call. Your Candid Health MCP server gets session-scoped tokens with zero credentials stored on your infra.
Agent Auth →Security
Every Candid Health tool response scanned for prompt injection in milliseconds — 88.7% accuracy, all running on CPU.
Prompt Injection Defense →Performance
Free up to 96% of your agent's context window to enhance reasoning and reduce cost, on every Candid Health call.
Tools Discovery →A Candid Health MCP server lets AI agents read and write Candid Health data through the Model Context Protocol — Anthropic's open standard for connecting LLMs to external tools. StackOne's Candid Health MCP server ships with 45 pre-built actions, fully extensible via the Connector Builder — plus managed authentication, prompt injection defense, observability, and agent execution runtime. Connect it from MCP clients like Claude Desktop, Claude Code, Cursor, Goose, and VS Code, or from agent frameworks like OpenAI Agents SDK, LangChain, and Vercel AI SDK.
Every action from Candid Health's API, ready for your agent. Create, read, update, and delete — scoped to exactly what you need.
Create a pre-encounter appointment for a patient with scheduling and eligibility metadata.
Retrieve a single appointment by its Candid appointment ID.
Update an appointment using optimistic concurrency via the next version number.
Create a charge capture holding pre-billing charge data before encounter submission.
List charge captures filtered by patient, bundle, status, service dates, or provider.
Retrieve a single charge capture by its Candid charge capture ID.
Update charge capture data or bundle assignment before aggregation.
Create an insurance coverage record for a patient with payer, plan, subscriber, and benefits data.
List coverages filtered by patient, payer, and status with cursor pagination.
Retrieve a single coverage record by its Candid coverage ID.
Update a coverage record using optimistic concurrency via the next version number.
Attach an ICD diagnosis code to an existing encounter.
Update an existing diagnosis code, code type, name, or POA indicator.
Create a professional encounter (claim) with patient, providers, diagnoses, and service lines.
List professional encounters filtered by status, dates, patient, provider, payer, and more.
Retrieve a single encounter (claim) with its providers, diagnoses, and service lines.
Update fields on an existing encounter (claim), including status, providers, and billing metadata.
Post an insurance refund with allocations to service lines, claims, or billing provider unattributed cash.
List insurance refunds filtered by payer, service line, claim, or billing provider.
Retrieve a single insurance refund and its allocations by refund ID.
Update an insurance refund's amount, timestamp, notes, or allocations.
Create a pre-encounter patient with demographics, contacts, guarantors, and other administrative data.
Search patients by MRN or other identifier fields.
List patients with cursor pagination, sort, filters, and free-text search.
Retrieve a single patient by their Candid patient ID.
Update a patient record using optimistic concurrency via the next version number.
Retrieve a single payer from Candid's canonical payer list by its UUID.
List all payers on Candid's canonical payer list with optional name search.
Add a service line (CPT/HCPCS code) with charges and diagnosis pointers to an encounter.
Update an existing service line's code, modifiers, units, charge, or diagnosis pointers.
Remove a service line from its encounter by service line ID.
Create a billing operations task tied to an encounter with type, category, and assignee.
List billing operations tasks filtered by status, category, assignee, or encounter.
Retrieve a single task with its notes, status history, and encounter reference.
Update a task's status, assignee, notes, or blocking reasons.
Post one or more write-offs applied to a service line, claim, or billing provider.
List write-offs filtered by payer, claim, service line, billing provider, or account type.
Retrieve a single write-off record by its Candid write-off ID.
List appointment visits within a date range with pagination and sorting.
Fetch CSV export files of claim version changes within a date window.
Retrieve a single insurance adjudication (ERA/EOB posting) by its Candid ID.
Submit a real-time 270/271 eligibility check to a payer via Stedi.
Submit a batch of eligibility checks to payers for asynchronous processing.
Deactivate (soft-delete) a patient using optimistic concurrency via the next version number.
Reverse a previously posted write-off, restoring the balance on its target.
One endpoint. Any framework. Your agent is talking to Candid Health in under 10 lines of code.
Agent Frameworks
{
"mcpServers": {
"stackone": {
"command": "npx",
"args": [
"-y",
"mcp-remote@latest",
"https://api.stackone.com/mcp?x-account-id=<account_id>",
"--header",
"Authorization: Basic <YOUR_BASE64_TOKEN>"
]
}
}
}Anthropic's code_execution processes data already in context. Custom MCP code mode keeps raw tool responses in a sandbox. 14K tokens vs 500.
11 min
Benchmarking BM25, TF-IDF, and hybrid search for MCP tool discovery across 916 tools. The 80/20 TF-IDF/BM25 hybrid hits 21% Top-1 accuracy in under 1ms.
10 min
MCP tools that read emails, CRM records, and tickets are indirect prompt injection vectors. Here's how we built a two-tier defense that scans tool results in ~11ms.
12 min
MCP vs A2A: what each protocol standardizes, how they differ, their shared security risks including indirect prompt injection, and when to use one, both, or a hybrid architecture.
12 min
MCP wraps APIs, it doesn't replace them. After building 200+ connectors that serve both, here's when each approach wins.
14 min read
origin_owner_id.All the tools you need to build and scale AI agent integrations, with best-in-class connectivity, execution, and security.