VapiSAFE
Vapi MCP Server
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
Build AI voice assistants and phone agents with Vapi using the Model Context Protocol.
Setup
The MCP server requires a Vapi API key. Get one from the Vapi dashboard.
Local Server (stdio)
Configure any MCP client that supports local stdio servers to run:
npx -y @vapi-ai/mcp-server
Set VAPI_TOKEN in the server environment. MCP client configuration formats vary, but the server definition generally looks like this:
{
"mcpServers": {
"vapi": {
"command": "npx",
"args": ["-y", "@vapi-ai/mcp-server"],
"env": {
"VAPI_TOKEN": ""
}
}
}
}Remote Server (Streamable HTTP)
Clients that support remote MCP servers can connect directly:
- URL:
https://mcp.vapi.ai/mcp - Header:
Authorization: Bearer your_vapi_api_key_here
Client-Specific Examples
Claude Code
claude mcp add -e VAPI_TOKEN=your_vapi_token vapi -- npx -y @vapi-ai/mcp-server
Claude Desktop
Use the local server configuration above in the Claude Desktop configuration file. To connect to the hosted server through an stdio bridge instead:
{
"mcpServers": {
"vapi": {
"command": "npx",
"args": [
"mcp-remote",
"https://mcp.vapi.ai/mcp",
"--header",
"Authorization: Bearer ${VAPI_TOKEN}"
],
"env": {
"VAPI_TOKEN": ""
}
}
}
}Optional Agent Skill
The skill directory contains reusable instructions for AI coding agents that support Agent Skills. Install it using your host's skill installation process.
For Claude Code:
mkdir -p ~/.claude/skills/vapi curl -o ~/.claude/skills/vapi/SKILL.md https://raw.githubusercontent.com/VapiAI/mcp-server/main/skill/SKILL.md
Example Usage
Create a Voice Assistant
Ask your MCP-enabled agent:
66ac59af48cbOBSERVED · 2026-10-08Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add mcp-server --env VAPI_TOKEN=${VAPI_TOKEN} -- npx -y @vapi-ai/[email protected]{
"mcpServers": {
"mcp-server": {
"command": "npx",
"args": [
"-y",
"@vapi-ai/[email protected]"
],
"env": {
"VAPI_TOKEN": "${VAPI_TOKEN}"
}
}
}
}Exposed tools (14)
9 read · 5 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
compatibility_tool | read | Compatibility tool |
create_assistant | write | Creates a new Vapi assistant |
create_call | write | Creates a outbound call |
create_tool | write | Creates a new Vapi tool |
get_assistant | read | Gets a Vapi assistant by ID |
get_call | read | Gets details of a specific call |
get_phone_number | read | Gets details of a specific phone number |
get_tool | read | Gets details of a specific tool |
list_assistants | read | |
list_calls | read | |
list_phone_numbers | read | |
list_tools | read | |
update_assistant | write | Updates an existing Vapi assistant |
update_tool | write | Updates an existing Vapi tool |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (4)
new URL(`../../${relativePath}`, import.meta.url),readFileSync(new URL('../../package.json', import.meta.url), 'utf8')readFileSync(new URL('../../package.json', import.meta.url), 'utf8')@vapi-ai/server-sdk, dotenv, zod, zod-to-json-schema, @types/jest, @types/node, jest, shx
Gates applied: no_behavioural_pass.
66ac59af48cbfull audit observations/trust-audit/mcp-server/vapiai__vapi.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-08 | 66ac59af48cb | SAFE | B | 89 | first audit |
Questions
What is the Vapi MCP server?
Vapi MCP Server
What tools does Vapi expose?
14 in total: 9 read-only, 5 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is Vapi safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.
What credentials does Vapi need?
It reads VAPI_TOKEN from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does Vapi run?
It speaks stdio and streamable-http, so it runs as a local process your client starts. It is published on npm as @vapi-ai/mcp-server at 0.0.11.
How current is this page?
The grade is for one exact copy of the source (66ac59af48cb), read on 2026-10-08. The repository is watched and re-audited when it changes.