Gives agents access to AgentOps observability and trace data for debugging AI agent runs.
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add agentops --env AGENTOPS_API_KEY='<AGENTOPS_API_KEY>' -- uv --directory path/to/your/main/directory run server.py~/.codex/config.toml
[mcp_servers.agentops]
command = "uv"
args = ["--directory", "path/to/your/main/directory", "run", "server.py"]
env = { AGENTOPS_API_KEY = "<AGENTOPS_API_KEY>" }
# codex mcp add agentops --env AGENTOPS_API_KEY='<AGENTOPS_API_KEY>' -- uv --directory path/to/your/main/directory run server.py~/.cursor/mcp.json
{
"mcpServers": {
"agentops": {
"command": "uv",
"args": [
"--directory",
"path/to/your/main/directory",
"run",
"server.py"
],
"env": {
"AGENTOPS_API_KEY": "<AGENTOPS_API_KEY>"
}
}
}
}.vscode/mcp.json
{
"servers": {
"agentops": {
"type": "stdio",
"command": "uv",
"args": [
"--directory",
"path/to/your/main/directory",
"run",
"server.py"
],
"env": {
"AGENTOPS_API_KEY": "<AGENTOPS_API_KEY>"
}
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"agentops": {
"command": "uv",
"args": [
"--directory",
"path/to/your/main/directory",
"run",
"server.py"
],
"env": {
"AGENTOPS_API_KEY": "<AGENTOPS_API_KEY>"
}
}
}
}claude_desktop_config.json
{
"mcpServers": {
"agentops": {
"command": "uv",
"args": [
"--directory",
"path/to/your/main/directory",
"run",
"server.py"
],
"env": {
"AGENTOPS_API_KEY": "<AGENTOPS_API_KEY>"
}
}
}
}.trae/mcp.json
{
"mcpServers": {
"agentops": {
"command": "uv",
"args": [
"--directory",
"path/to/your/main/directory",
"run",
"server.py"
],
"env": {
"AGENTOPS_API_KEY": "<AGENTOPS_API_KEY>"
}
}
}
}Environment variables
AGENTOPS_API_KEY
Replace the placeholders with your own credentials and keep them out of version control.