Agentsim

dev.agentsimPublished by agentsim.devOfficial publisherv0.33.3Updated Oct 24, 2026

Control plane for agent auth walls. Connector 0 handles SMS OTP with US test numbers.

VerifiedStreamable HTTPWeb executableDeveloper Tools Communication & Collaboration Security & Monitoring

Server details

Identifier: dev.agentsim/mcp
Transport: Streamable HTTP
Runtime: Web executable
Trust: Verified
Published: Sep 16, 2026

Security note

MCP servers receive tool arguments from a conversation and may perform external actions. Review what a server can do before you enable it, and give it only the access it needs.

The registry verified that this server's namespace belongs to agentsim.dev, which also runs its endpoint. That identifies the publisher; it does not make the server's tools safe.

SourceWeft's checks cover whether the listing is complete and the endpoint answered; they do not review the server's code or what its tools do.

Add a badge to your README

Show readers this MCP server is listed on SourceWeft. The badge links to this page.

[![Listed on SourceWeft](https://sourceweft.com/badge/mcp/dev.agentsim%2Fmcp.svg)](https://sourceweft.com/mcp/dev.agentsim%2Fmcp?utm_source=readme&utm_medium=badge)

Overview

AI-generated overview

Lets an AI coding assistant open authentication challenges for owned services and wait for SMS OTP, email OTP, magic link, or WebAuthn verdicts.

What you can do: Exposes AgentSIM challenge tools to MCP-compatible coding assistants. open_challenge opens an authentication challenge session for a required owned service_url and accepts a channel such as sms_otp, email_otp, magic_link, or webauthn_required, returning a session ID and identifier. wait_for_verdict long-polls for the verdict and returns a structured outcome including otp_code, magic_link, or webauthn_required. Supporting tools include identify_agent, get_messages, release_number, and list_numbers, plus legacy aliases provision_number and wait_for_otp.
When to use it: Useful when an assistant needs to complete a phone or email sign-in flow against a service you own during development or testing, for example driving an SMS OTP challenge end to end. It fits coding assistants such as Codex CLI, Claude Code, Cursor, and Windsurf that support MCP. Not intended for services you do not own or for non-US SMS numbers.
Requirements: Remote streamable HTTP endpoint at or a local stdio process via uvx agentsim-mcp. Authentication is required: AGENTSIM_API_KEY in the process environment for stdio, or an x-api-key header on every hosted HTTP request. A key is obtained from the AgentSIM console. Live SMS is limited to US numbers and requires an owned public HTTPS service with exact-origin verification and an allow policy.

Before you install

Requires an API key (AGENTSIM_API_KEY for stdio, x-api-key header for hosted HTTP) that is caller-scoped and forwarded by the server. Live SMS sessions are metered: Hobby includes 10 live US SMS sessions per UTC month with one active live session per account, and Builder is $99/month for 50 sessions, so usage can cost money. Reading codes with get_messages consumes them. Live SMS requires exact-origin verification and an allow policy, and policy denial fails on open.

Installation

In SourceWeft

Open Agentsim in the dashboard and add it to a workspace.
Connect the required credentials. They are stored privately and never shown on public pages.
Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/dev.agentsim%2Fmcp/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config. The client completes OAuth sign-in on first connect.

{
  "mcpServers": {
    "agentsim": {
      "type": "http",
      "url": "https://mcp.agentsim.dev/mcp"
    }
  }
}

README

agentsim-mcp

Availability: Live SMS is self-serve for owned public HTTPS services after exact-origin verification and an allow policy. Hobby includes 10 live US SMS sessions per UTC month, with one active live session per account and no card required. Builder is $99/month and includes 50 live US SMS sessions. Checkout shows tax and renewal terms before payment, and existing agreements keep their terms. Check current access and channel support before using live examples.

MCP server that exposes AgentSIM challenge tools to AI coding assistants: Codex CLI, Claude Code, Cursor, Windsurf, and any other MCP-compatible host. Primary tools: open_challenge, wait_for_verdict. Aliases: provision_number, wait_for_otp.

Setup

For Codex CLI, follow Connect Codex. It covers key handling, the SMS wait timeout, and browser access. Then use the complete phone sign-in prompt.

Claude Code

Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json:

json

{  "mcpServers": {    "agentsim": {      "command": "uvx",      "args": ["agentsim-mcp"],      "env": {        "AGENTSIM_API_KEY": "asm_live_xxx"      }    }  }}

{  "mcpServers": {    "agentsim": {      "command": "uvx",      "args": ["agentsim-mcp"],      "env": {        "AGENTSIM_API_KEY": "asm_live_xxx"      }    }  }}

Cursor / Windsurf

Add agentsim-mcp as a stdio MCP server with AGENTSIM_API_KEY in the environment config.

Remote (no install)

Connect directly to the hosted MCP server without installing anything locally:

json

{  "mcpServers": {    "agentsim": {      "type": "streamable-http",      "url": "https://mcp.agentsim.dev/mcp",      "headers": {        "x-api-key": "asm_live_..."      }    }  }}

{  "mcpServers": {    "agentsim": {      "type": "streamable-http",      "url": "https://mcp.agentsim.dev/mcp",      "headers": {        "x-api-key": "asm_live_..."      }    }  }}

The hosted endpoint uses stateless HTTP and does not issue mcp-session-id headers. The session_id returned by AgentSIM tools identifies a challenge and remains valid across tool calls.

Tools

Control-Plane Nouns (Recommended)

Tool	Description
`identify_agent`	Add a stable agent ID to the console without opening a challenge, sending SMS, or using allowance
`open_challenge`	Open an authentication challenge session for a required owned `service_url`; accepts channel (sms_otp \| email_otp \| magic_link \| webauthn_required), returns session ID + identifier
`wait_for_verdict`	Long-poll for the challenge verdict — returns structured outcome including otp_code, magic_link, or webauthn_required; policy denial fails on open
`get_messages`	Read SMS metadata and parsed codes; reading codes consumes them
`release_number`	Release a session early (allocation returned to pool)
`list_numbers`	List all active sessions for this account

Legacy Aliases (Backward Compatibility)

Tool	Maps to	Description
`provision_number`	`open_challenge(channel=sms_otp)`	Provision a temporary programmable US number for SMS OTP — kept for backward compatibility
`wait_for_otp`	`wait_for_verdict`	Legacy wait with automatic extension after timeout by default; number may be unchanged

Auth

For stdio, set AGENTSIM_API_KEY in the process environment. For hosted HTTP, every request must send its own x-api-key; the server validates and forwards that caller-scoped key. Get a key at console.agentsim.dev.

Supported Countries

Source: README.md at commit aa484e9

Tools

This server asks for a credential such as an API key, so SourceWeft could not read its tools without one and has not indexed them.

Version history

v0.33.3LatestSep 29, 2026
v0.33.2Sep 29, 2026
v0.33.1Sep 29, 2026
v0.33.0Sep 28, 2026
v0.32.2Sep 27, 2026
v0.32.0Sep 27, 2026
v0.31.0Sep 25, 2026
v0.30.0Sep 24, 2026
v0.29.1Sep 16, 2026

Score

85% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

Reached in the last 30 daysRequired25 pts
Auth declaredRequired10 pts
Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
DescriptionRequired15 pts
Logo10 pts
Documentation15 pts
License10 pts