Replai MCP
Replai's official remote MCP server enables AI agents to manage AI customer support agents, including creating drafts, importing knowledge sources, running tests, publishing updates, managing the inbox, and integrating with engineering tools like Linear.
Endpoint and access
OAuth sign-in: the first time you connect, your client opens the vendor's own authorization prompt in a browser.
https://api.replai.tw/mcpAbout Replai MCP
Replai MCP is an official remote server by AIthentic LLC designed for productivity and knowledge management. It enables AI agents to interact with AI customer support systems. Key capabilities include creating and editing agent drafts, importing knowledge from websites and files, and managing inbox conversations.
The server supports operational tasks such as running tests to review knowledge gaps, publishing updates, and integrating with engineering tools like Linear. These functions allow agents to maintain and improve support workflows autonomously.
Access requires OAuth sign-in through a browser on the same computer. Users must have a Replai account and sign in via Google at replai.tw. An AI tool supporting remote MCP servers with OAuth is also required.
What you can do with it
- Create and edit AI agent drafts
- Import knowledge from websites and files
- Run tests and review knowledge gaps
- Manage inbox conversations and replies
Source and maintenance
Tools
Reported by the attributed source; not independently observed by Koskamo.
Maintenance history
Public check and source-refresh events. Editorial and ownership records remain private.
- · registry · passed · metadata changed · Source metadata updated
- · backfill · recorded · Pre-expansion baseline; unobserved rich metadata remains unknown
Connect from your client
Each snippet registers the same endpoint; pick the client you use.
Claude Code
claude mcp add --transport http replai-mcp https://api.replai.tw/mcpCodex
codex mcp add replai-mcp --url https://api.replai.tw/mcpCursor
{
"mcpServers": {
"replai-mcp": {
"url": "https://api.replai.tw/mcp"
}
}
}Cline
{
"mcpServers": {
"replai-mcp": {
"type": "streamableHttp",
"url": "https://api.replai.tw/mcp",
"disabled": false
}
}
}VS Code
{
"servers": {
"replai-mcp": {
"type": "http",
"url": "https://api.replai.tw/mcp"
}
}
}Gemini CLI
gemini mcp add --transport http replai-mcp https://api.replai.tw/mcpClaude Desktop
{
"mcpServers": {
"replai-mcp": {
"url": "https://api.replai.tw/mcp"
}
}
}Windsurf
{
"mcpServers": {
"replai-mcp": {
"serverUrl": "https://api.replai.tw/mcp"
}
}
}Let your agent find its tools
Find servers like Replai MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.
Other productivity and knowledge servers
Frequently asked questions
- Can I use Replai MCP without a Replai account?
- No. A Replai account is a documented requirement. You must sign in with Google at replai.tw to authenticate and access the server's tools.
- What tasks does the server support?
- The server supports creating and editing AI agent drafts, importing knowledge from websites and files, managing inbox conversations, running tests, publishing updates, and integrating with Linear.
- Do I need a browser to connect?
- Yes. The first connection opens the vendor's authorization prompt in a browser. You must use a browser on the same computer to approve OAuth access.
- Does the server integrate with engineering tools?
- Yes. The server supports integration with engineering tools like Linear. This allows agents to manage updates and workflows within those external systems.
Do you run Replai MCP?
Prove ownership with a DNS record, a well-known file, or your Official MCP Registry entry, and your edits publish immediately. Proofs are re-checked weekly; admins can still hide or remove a listing.