Karbon MCP Server
The Karbon MCP Server connects AI assistants to the Karbon practice management platform, enabling natural language interaction with your firm's data across these capabilities:
Contacts & Clients
Contacts: List, search, create, and update person contacts; look up by Karbon key or external ID
Organizations: List, search, create, and update organization contacts
Client Groups: List, search, create, and update client groups (households/entities)
Business Cards: Get and update contact details (email, phone, address, websites)
Work Management
Work Items: List, search, create, and update jobs/engagements; filter by status, assignee, due date, etc.
Estimate Summary: View budgeted vs. actual time and value for budget tracking
Work Templates: List and retrieve templates for use when creating work items
Work Schedules: Get, create, and update recurring schedules that auto-generate work items
Time & Billing
Timesheets & Time Entries: List and retrieve timesheets and individual time entries (read-only)
Invoices: List and retrieve invoices with optional line items and payment data (read-only)
Payments: List and retrieve payments; optionally create, delete, and reverse manual payments (requires opt-in via
KARBON_ALLOW_PAYMENT_WRITES=true)
Notes & Communication
Notes & Comments: Retrieve and create notes linked to work items, contacts, or client groups; retrieve comments from timelines
Files
List files attached to work items, contacts, or organizations; download files to disk; upload files to entities
Custom Fields
List definitions, get/set values on entities, and create or delete custom field definitions tenant-wide
Teams & Users
Teams: List, retrieve, and manage team membership (add/remove members)
Users: List and retrieve Karbon user accounts
Account & Configuration
Retrieve tenant settings: valid work types, statuses, and contact types
Query Power
All list operations support OData filtering (
$filter), ordering ($orderby), and pagination ($top/$skip)Read-only mode available to prevent any data modifications
Remote HTTP mode available for connecting ChatGPT, claude.ai, and other remote MCP clients
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Karbon MCP Server show my in-progress work items due this month"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Karbon MCP Server
🚧 Early release — this project is brand new and we're actively making updates over the next few days. Things may change quickly; pin a version if you need stability, and check back soon.
An unofficial Model Context Protocol server for Karbon practice management. Connect Claude Code, Claude Desktop, GitHub Copilot, or any MCP-compatible client to your Karbon instance and work with contacts, work items, time, invoices, payments, files, and more using natural language.
Disclaimer: This is a community project and is not affiliated with or endorsed by Karbon. Use at your own risk.
Features
Area | Tools |
Contacts |
|
Organizations |
|
Client groups |
|
Work items |
|
Notes & comments |
|
Timesheets |
|
Invoices |
|
Payments |
|
Business cards |
|
Custom fields |
|
Files |
|
Teams |
|
Work schedules |
|
Users |
|
Account |
|
get_contact, get_organization, and get_client_group can also look records up by your own UserDefinedIdentifier instead of the Karbon key — handy if you sync IDs from your tax software.
All list tools support OData filtering ($filter), ordering ($orderby), and pagination ($top/$skip) — so you can ask things like "show my in-progress work items due this month" and the AI can express that as a precise query.
Related MCP server: Kanta MCP Server
Prerequisites
Node.js 18+
Karbon API credentials. In Karbon go to Settings → Connected Apps → API Applications and copy your Bearer Token and Access Key. (Karbon help article)
Setup
Step-by-step walkthroughs for Claude Desktop, claude.ai/Cowork, and ChatGPT (written for non-developers) are in docs/client-guides.md.
Claude Desktop: one-click bundle
Download karbon-mcp-server.mcpb from the latest release and double-click it. Claude Desktop opens a form for your two Karbon keys (stored securely by your OS) plus a read-only toggle — no terminal needed. Walkthrough.
One-line install (nothing to download first)
No Node.js? These check for it (installing via winget/Homebrew where possible), then run the setup wizard straight from npm — no repo clone, no manual download:
# Windows (PowerShell)
irm https://raw.githubusercontent.com/Mad-Man-Dan/karbon-mcp-server/main/scripts/install.ps1 | iex# macOS / Linux
curl -fsSL https://raw.githubusercontent.com/Mad-Man-Dan/karbon-mcp-server/main/scripts/install.sh | bashQuickest with Node.js installed: automatic setup
npx -y karbon-mcp-server setupSetup asks a few quick questions — project or user-level install, your API keys (skippable), and whether to enable read-only mode — then does everything else automatically:
This project (the default): writes
.mcp.json(Claude Code), plus.vscode/mcp.jsonand.cursor/mcp.jsonif those folders exist in the project.User account: detects which MCP clients are installed on your machine — Claude Code, Codex, Claude Desktop, VS Code, Cursor — and adds the server to each (for Claude Code and Codex it uses their own
mcp addCLI).
It then offers to take your two Karbon keys (Enter skips either one):
Input is masked — you see a
*per character, then a first-4/last-4 preview (like1a2b****5c6d, the same style Karbon's site uses) so you can confirm what was entered, with a gentle warning if the format looks off (the Bearer Token is a GUID; the Access Key starts withey...).Keys are written only to your MCP client's local config file on your machine — the same file a manual setup would use. They are never sent anywhere else.
If you skip, placeholder values are written and setup tells you exactly which file(s) to edit later.
It backs up every existing config file before touching it, merges rather than overwrites, and never replaces real credentials with placeholders (re-running and skipping the prompts keeps your existing keys). If you go project-level, gitignore the config files — your keys live in them in plain text.
Options:
npx -y karbon-mcp-server setup --scope project # skip the question
npx -y karbon-mcp-server setup --scope user
npx -y karbon-mcp-server setup --read-only # disable all write tools
npx -y karbon-mcp-server setup --client vscode,cursor # only these clients
# ids: claude-code, codex, claude-desktop, vscode, cursor
npx -y karbon-mcp-server setup --print # just print the JSONOne-click installs
The Install in VS Code / Cursor badges above create the server entry for you with placeholder credentials. After installing, replace YOUR_BEARER_TOKEN and YOUR_ACCESS_KEY with your real keys — in VS Code via the Command Palette → MCP: Open User Configuration, in Cursor by editing ~/.cursor/mcp.json.
If you run the server before replacing the placeholders, it starts fine and every tool responds with setup instructions instead of failing cryptically — so you can even ask your AI assistant "why isn't Karbon working?" and it will tell you what to do.
Prefer to configure things by hand? Each client is below.
Claude Code
claude mcp add karbon \
--env KARBON_BEARER_TOKEN=your-bearer-token \
--env KARBON_ACCESS_KEY=your-access-key \
-- npx -y karbon-mcp-serverClaude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"karbon": {
"command": "npx",
"args": ["-y", "karbon-mcp-server"],
"env": {
"KARBON_BEARER_TOKEN": "your-bearer-token",
"KARBON_ACCESS_KEY": "your-access-key"
}
}
}
}VS Code (GitHub Copilot)
Add to .vscode/mcp.json in your workspace (or your user mcp.json):
{
"servers": {
"karbon": {
"command": "npx",
"args": ["-y", "karbon-mcp-server"],
"env": {
"KARBON_BEARER_TOKEN": "your-bearer-token",
"KARBON_ACCESS_KEY": "your-access-key"
}
}
}
}Cursor
Add to ~/.cursor/mcp.json (or use the install badge above):
{
"mcpServers": {
"karbon": {
"command": "npx",
"args": ["-y", "karbon-mcp-server"],
"env": {
"KARBON_BEARER_TOKEN": "your-bearer-token",
"KARBON_ACCESS_KEY": "your-access-key"
}
}
}
}Codex (OpenAI Codex CLI)
codex mcp add karbon \
--env KARBON_BEARER_TOKEN=your-bearer-token \
--env KARBON_ACCESS_KEY=your-access-key \
-- npx -y karbon-mcp-serverOr add to ~/.codex/config.toml:
[mcp_servers.karbon]
command = "npx"
args = ["-y", "karbon-mcp-server"]
[mcp_servers.karbon.env]
KARBON_BEARER_TOKEN = "your-bearer-token"
KARBON_ACCESS_KEY = "your-access-key"ChatGPT, claude.ai chat & Claude Cowork (remote connectors)
These clients can't launch local servers — they only connect to a remote MCP server over HTTPS. This package includes a streamable-HTTP mode for exactly that, so a firm can self-host it and keep their Karbon keys in their own infrastructure:
KARBON_BEARER_TOKEN=... KARBON_ACCESS_KEY=... KARBON_HTTP_SECRET=some-long-random-string \
npx -y karbon-mcp-server --http --port 8787This serves the MCP endpoint at /mcp/<your-secret> (plus a /health check). Expose it over HTTPS however you like — the quickest free option is a Cloudflare Tunnel:
cloudflared tunnel --url http://localhost:8787Then add the URL (e.g. https://your-tunnel.trycloudflare.com/mcp/your-secret) as:
claude.ai / Cowork: Settings → Connectors → Add custom connector (Pro/Max/Team/Enterprise plans)
ChatGPT: Settings → Connectors (requires developer mode for full tool access)
Security notes: anyone with the full URL has the same Karbon access as your API keys, so use a long random
KARBON_HTTP_SECRET, always front it with HTTPS, and considerKARBON_READ_ONLY=truefor remote use. There is deliberately no shared/hosted instance of this server — your keys should only ever live on infrastructure you control.
Configuration
Environment variable | Required | Description |
| Yes | Bearer token from Karbon Connected Apps |
| Yes | Access key (JWT) from Karbon Connected Apps |
| No | Set to |
| No | Set to |
| No | Override the API base URL (default |
| No | HTTP mode only: secret path segment for the endpoint ( |
| No | HTTP mode only: port to listen on (default 8787; |
Read-only mode
If you want a guarantee that the AI can never modify your Karbon data, set KARBON_READ_ONLY=true (or pass --read-only). Write tools are not registered at all in this mode, so the AI never sees them. This also disables download_file — it doesn't change Karbon, but it writes to the local disk, and read-only means no writes anywhere. The setup wizard asks about this.
Changing it later: setup always writes the toggle explicitly ("KARBON_READ_ONLY": "true" or "false"), so the switch is right there in your config file — flip the value and restart the client. Write tools appear or disappear accordingly (in read-only mode they aren't registered at all, so the AI never sees them). Alternatively, re-run setup and skip the key prompts (press Enter) — your saved keys are kept, and your new read-only answer is applied. One caveat: entries managed by the claude / codex CLIs aren't touched by a re-run — for those, run claude mcp remove karbon (or codex mcp remove karbon) first, then re-run setup.
Payment writes (extra opt-in)
Recording, deleting, or reversing payments touches financial records, so those three tools sit behind their own switch — even with writes otherwise enabled they are not registered until you set KARBON_ALLOW_PAYMENT_WRITES=true (or pass --allow-payment-writes). Setup writes the toggle as "false" so it's easy to find and flip in your config file; re-running setup never resets a value you changed by hand. Read-only mode always wins: if KARBON_READ_ONLY=true, payment writes stay off regardless.
A note on credentials
Whether you type your keys into the setup wizard or paste them into a config file by hand, they end up in exactly one place: your own MCP client's local config file — the same file a fully manual setup would use. Nothing in this project transmits your keys anywhere else. Like every stdio MCP server, the keys live in that file in plain text, so treat it like a password: it grants the same access to your Karbon account as your keys do. If a key is ever exposed, regenerate it in Karbon under Settings → Connected Apps.
Example prompts
"Find the contact record for Jane Smith and show her business card details."
"Update Jane Smith's email address and phone number."
"List all in-progress work items assigned to me, ordered by due date."
"Are we over budget on the Hendersons' tax return?"
"Which invoices are still awaiting payment, and for how much?"
"Create a work item for Acme Corp's 2025 tax return, assigned to alex@firm.com, starting next Monday."
"Add a note to the Smith work item summarizing today's client call."
"How many hours were logged against client work last week?"
"What files are attached to the Miller onboarding work item?"
Development
git clone https://github.com/Mad-Man-Dan/karbon-mcp-server.git
cd karbon-mcp-server
npm install
npm test # builds, then runs the smoke tests (tool matrix across flag combos)Releases are automated: pushing a v* tag runs CI, publishes to npm via trusted publishing (OIDC — no tokens), and updates the MCP Registry listing.
Test locally with the MCP Inspector:
npx @modelcontextprotocol/inspector \
-e KARBON_BEARER_TOKEN=... -e KARBON_ACCESS_KEY=... \
node dist/index.jsThe Karbon API is documented at developers.karbonhq.com with an OpenAPI spec at karbonhq/karbon-api-reference.
Contributing
Contributions welcome — open an issue or PR.
License
Available Tools
50 toolsadd_team_membersAdd team membersA
Add one or more users to a team by their UserKeys (find them with list_users). Users already on the team are skipped. Team membership can affect work visibility and assignment — confirm with the user before changing it.
Inputs: teamKey (required), UserKeys (required).
create_client_groupCreate client groupB
Create a new client group in Karbon. Members are existing contacts and/or organizations.
Inputs: FullName (required), Members, ClientOwner, ContactType, +5 more.
create_contactCreate contactC
Create a new person contact in Karbon.
Inputs: LastName (required), FirstName (required), MiddleName, Salutation, +7 more.
create_custom_fieldCreate custom field definitionA
Create a new custom field definition for the whole Karbon tenant. This changes the firm's shared configuration, not just one record — state the field name, type, and visibility to the user and get their explicit confirmation before calling.
Inputs: Name (required), Type (required), ListOptions, IsVisibleToContacts, +1 more.
create_noteCreate noteA
Create a note in Karbon, optionally linked to work items, contacts, organizations, or client groups via timelines. Notes linked to a work item also appear on the client's timeline.
Inputs: Body (required), Subject (required), AuthorEmailAddress (required), DueDate, +3 more.
create_organizationCreate organizationB
Create a new organization contact in Karbon.
Inputs: FullName (required), ClientOwner, ContactType, ClientManager, +2 more.
create_work_itemCreate work itemA
Create a new work item in Karbon. WorkType and status values are tenant-specific — call get_tenant_settings first if unsure.
Inputs: Title (required), ClientKey (required), StartDate (required), ClientType (required), +9 more.
create_work_scheduleCreate work scheduleA
Create a recurring work schedule from an existing work item. The schedule will automatically generate new work items on the recurrence you set — confirm the recurrence details with the user before calling. Note: RecurrenceFrequency cannot be changed after creation.
Inputs: ScheduleStartDate (required), RecurrenceFrequency (required), ScheduleDueDateMethod (required), CreatedFromWorkItemKey (required), +12 more.
delete_custom_fieldDelete custom field definitionA
DESTRUCTIVE AND IRREVERSIBLE: deleting a custom field definition permanently removes the field AND every value stored in it across all contacts, organizations, and client groups in the tenant. Never call this without first telling the user exactly which field will be deleted (by name, via list_custom_fields) and receiving their explicit confirmation for that specific field.
Inputs: customFieldDefinitionKey (required).
download_fileDownload a fileA
Download a Karbon file to a local path on the machine running this MCP server. Get the DownloadUrl from list_entity_files first — download links expire after 15 minutes, so re-list if the download fails with an auth error. Ask the user where to save before writing outside a temp directory, and never overwrite a file the user didn't ask you to replace.
Inputs: savePath (required), downloadUrl (required).
get_business_cardGet business cardA
Get a single business card by BusinessCardKey. Business cards hold the contact details (email, phone, address, websites) of a contact, organization, or client group. Find the key by expanding BusinessCards on get_contact / get_organization / get_client_group.
Inputs: businessCardKey (required).
get_client_groupGet client groupA
Get a single client group by ClientGroupKey — or by your own UserDefinedIdentifier — optionally expanding related data (BusinessCard, ClientTeam). Includes the group's members (contacts and organizations). Provide exactly one of clientGroupKey or userDefinedIdentifier.
Inputs: expand, clientGroupKey, userDefinedIdentifier.
get_commentGet commentA
Get a single comment by its CommentKey. Comments appear on work item and client timelines alongside notes.
Inputs: commentKey (required).
get_contactGet contactA
Get a single person contact by ContactKey — or by your own UserDefinedIdentifier — optionally expanding related data (BusinessCards, AccountingDetail, ClientTeam). Provide exactly one of contactKey or userDefinedIdentifier.
Inputs: expand, contactKey, userDefinedIdentifier.
get_custom_field_valuesGet custom field valuesA
Get the custom field values set on a specific entity (contact, organization, or client group) by its entity key.
Inputs: entityKey (required).
get_estimate_summaryGet estimate summaryA
Get the estimate summary for a work item (read-only): budgeted vs. actual time and value. Useful for 'are we over budget on this job' questions — pair with list_time_entries for detail.
Inputs: workItemKey (required).
get_invoiceGet invoiceA
Get a single invoice by InvoiceKey (read-only), optionally expanding related data (LineItems, Payments, Data).
Inputs: invoiceKey (required), expand.
get_noteGet noteB
Get a single note by its NoteID.
Inputs: noteId (required).
get_organizationGet organizationA
Get a single organization by OrganizationKey — or by your own UserDefinedIdentifier — optionally expanding related data (BusinessCards, AccountingDetail, ClientTeam, Contacts). Provide exactly one of organizationKey or userDefinedIdentifier.
Inputs: expand, organizationKey, userDefinedIdentifier.
get_paymentGet paymentA
Get a single payment by PaymentKey (read-only).
Inputs: paymentKey (required).
get_teamGet teamA
Get a single team by TeamKey, including its members (users and sub-teams).
Inputs: teamKey (required).
get_tenant_settingsGet tenant settingsA
Get this Karbon account's configuration: valid work statuses (secondary statuses), work types, and contact types. Call this before creating or updating work items or contacts with tenant-specific values.
Inputs: no named parameters.
get_time_entryGet time entryA
Get a single individual time entry by its IndividualTimeEntryKey (read-only).
Inputs: timeEntryKey (required).
get_timesheetGet timesheetA
Get a single timesheet by TimesheetKey, optionally expanding TimeEntries.
Inputs: timesheetKey (required), expandTimeEntries.
get_userGet userA
Get a single Karbon user by their UserId.
Inputs: userId (required).
get_work_itemGet work itemB
Get a single work item by WorkItemKey.
Inputs: workItemKey (required).
get_work_scheduleGet work scheduleA
Get a recurring work schedule by WorkScheduleKey — the recurrence rules that automatically generate work items.
Inputs: workScheduleKey (required).
get_work_templateGet work templateA
Get a single work template by WorkTemplateKey, including its task and section structure.
Inputs: workTemplateKey (required).
list_client_groupsList client groupsA
List or search client groups in Karbon. Supports OData filtering, e.g. filter: "contains(FullName,'Smith Family')" or "UserDefinedIdentifier eq 'ABC123'".
Inputs: top, skip, filter, orderby.
list_contactsList contactsA
List or search person contacts in Karbon. Supports OData filtering, e.g. filter: "contains(FullName,'Smith')" or "ContactType eq 'Client'".
Inputs: top, skip, filter, orderby.
list_custom_fieldsList custom field definitionsA
List all custom field definitions for the Karbon tenant — field names, types (Text, Number, Date, Boolean, Colleague, ListSingleSelect, ListMultipleSelect), list options, and which entities they apply to.
Inputs: no named parameters.
list_entity_filesList files on an entityA
List the files attached to a work item, contact, or organization — name, size, mime type, created date, and a DownloadUrl usable with download_file. Download links expire 15 minutes after this call.
Inputs: entityKey (required), entityType (required).
list_invoicesList invoicesA
List invoices in Karbon (read-only). Supports OData filtering, e.g. filter: "InvoiceStatus eq 'AwaitingPayment'" or "TotalAmountDue gt 0". Statuses: Approved, AwaitingPayment, Paid, Exported, Voided.
Inputs: top, skip, filter, orderby.
list_organizationsList organizationsA
List or search organization contacts in Karbon. Supports OData filtering, e.g. filter: "contains(FullName,'Acme')".
Inputs: top, skip, filter, orderby.
list_paymentsList paymentsA
List payments recorded against invoices in Karbon (read-only). Supports pagination only — the Payments endpoint does not accept OData filters.
Inputs: top, skip.
list_teamsList teamsA
List teams in Karbon (read-only). Supports OData filtering, e.g. filter: "contains(Name,'Tax')".
Inputs: top, skip, filter.
list_time_entriesList time entriesA
List individual time entries (read-only). Useful for reporting on time logged against clients or work items.
Inputs: top, skip, filter, orderby.
list_timesheetsList timesheetsA
List timesheets (read-only). Useful filters: "StartDate ge 2026-07-01", "UserKey eq '...'".
Inputs: top, skip, filter, orderby.
list_usersList usersB
List Karbon users (team members) on this account.
Inputs: top, skip, filter, orderby.
list_work_itemsList work itemsA
List or search work items (jobs/engagements) in Karbon. Useful filters: "PrimaryStatus eq 'InProgress'", "AssigneeEmailAddress eq 'user@firm.com'", "DueDate lt 2026-08-01". PrimaryStatus values: Planned, ReadyToStart, InProgress, Waiting, Completed.
Inputs: top, skip, filter, orderby.
list_work_templatesList work templatesA
List work templates available on this account. Use a WorkTemplateKey with create_work_item to create templated work.
Inputs: top, skip, filter, orderby.
remove_team_memberRemove team memberA
Remove a user from a team by their UserKey. Team membership can affect work visibility and assignment — confirm with the user before changing it.
Inputs: UserKey (required), teamKey (required).
set_custom_field_valuesSet custom field valuesA
Set custom field values on an entity (contact, organization, or client group). Use list_custom_fields for the available definitions and get_custom_field_values to see what's currently set. Values are always passed as an array of strings, even for single values.
Inputs: entityKey (required), CustomFieldValues (required).
update_business_cardUpdate business cardA
Update a business card (full replace via PUT) — this is how you change a contact's, organization's, or client group's email addresses, phone numbers, and addresses. Fetch the card first with get_business_card and send back the complete card with your changes, since omitted fields are cleared.
Inputs: EntityKey (required), EntityType (required), businessCardKey (required), WebSites, +10 more.
update_client_groupUpdate client groupA
Update fields on an existing client group (partial update via PATCH). Only include the fields you want to change.
Inputs: clientGroupKey (required), FullName, ClientOwner, ContactType, +5 more.
update_contactUpdate contactA
Update fields on an existing person contact (partial update via PATCH). Only include the fields you want to change.
Inputs: contactKey (required), LastName, FirstName, MiddleName, +7 more.
update_organizationUpdate organizationB
Update fields on an existing organization (partial update via PATCH). Only include the fields you want to change.
Inputs: organizationKey (required), FullName, ClientOwner, ContactType, +3 more.
update_work_itemUpdate work itemA
Update fields on an existing work item (partial update via PATCH). Only include the fields you want to change.
Inputs: workItemKey (required), Title, DueDate, WorkType, +4 more.
update_work_scheduleUpdate work scheduleA
Update a work schedule (partial via PATCH). The API only supports changing the end date and the assignee here. Note: changing AssigneeUserKey reassigns every not-yet-started work item the schedule has generated — confirm with the user first.
Inputs: workScheduleKey (required), AssigneeUserKey, ScheduleEndDate.
upload_fileUpload a fileA
Upload a file to Karbon and attach it to a work item, contact, organization, or client group (at least one key is required). SIZE LIMITS: keep uploads small — Karbon rejects oversized uploads, and this server reads the whole file into memory. Prefer filePath (a path on the machine running this MCP server) whenever possible; only use contentBase64 for small files (roughly under 1 MB), because base64 inflates data ~33% and large payloads can exceed the AI client's message limits or be truncated mid-transfer. If a file seems large (tens of MB or more), tell the user to upload it through Karbon's own UI instead. SECURITY: only upload files the user explicitly pointed you to. Never upload configuration files, credential stores, environment files, keys, or anything the user has not named — uploading puts the file's contents in the Karbon tenant.
Inputs: fileName, filePath, contactKey, workItemKey, +3 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 50 tools
Each tool targets a distinct resource and action (e.g., create_contact, list_invoices, delete_custom_field). There is no ambiguity as tool names clearly differentiate between different entities and operations.
All tools follow a consistent verb_noun pattern using snake_case (e.g., create_work_item, list_users, update_contact). No mixing of conventions or vague verbs.
50 tools is on the higher side but appropriate given the breadth of resources managed (contacts, work items, invoices, timesheets, files, etc.). Each tool serves a specific purpose, though the count could feel heavy for some users.
The tool set covers CRUD operations for most entities (contacts, organizations, work items, custom fields, etc.) with few notable gaps (e.g., no delete for contacts, work items, or notes). Overall, the surface is comprehensive for typical workflows.
Maintenance
Related MCP Connectors
Model Context Protocol server for the Apideck Unified API. Connect any MCP-compatible agent framework to 100+ accounting systems, HRIS platforms, file storage providers, and more through one integration. More information https://www.apideck.com/mcp-server
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to interact with yo…
Hosted MCP server for Cliniko — patients, appointments, availability, and invoices for AI agents.
Official Fibery MCP server: query and update your workspace from any AI client via OAuth.
Related MCP Servers
- AlicenseNot gradedqualityBmaintenanceA Model Context Protocol (MCP) server implementation that enables AI assistants to interact with Anytype's API through natural language, allowing users to manage their knowledge base through conversation.2,277 npm538MIT
- FlicenseNot gradedqualityDmaintenanceA Model Context Protocol server that enables AI assistants to interact with Kanta API for managing clients, users, persons, and accessing risk summaries and firm information.1-
- AlicenseNot gradedqualityDmaintenanceA Model Context Protocol (MCP) server that exposes Kanboard API functionality to Large Language Models (LLMs), enabling AI assistants to interact with Kanboard project management system.24 PyPI2MIT
- AlicenseNot gradedqualityBmaintenanceUnofficial MCP server for Akiflow that enables AI agents to manage tasks, projects, and calendar via the Model Context Protocol.194 npmMIT