Capture local AI agent traces and inspect failures, approvals, token usage, and run regressions through MCP.
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add trace -- uvx abhishekash-mcp-trace~/.codex/config.toml
[mcp_servers.trace]
command = "uvx"
args = ["abhishekash-mcp-trace"]
# codex mcp add trace -- uvx abhishekash-mcp-trace~/.cursor/mcp.json
{
"mcpServers": {
"trace": {
"command": "uvx",
"args": [
"abhishekash-mcp-trace"
]
}
}
}.vscode/mcp.json
{
"servers": {
"trace": {
"type": "stdio",
"command": "uvx",
"args": [
"abhishekash-mcp-trace"
]
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"trace": {
"command": "uvx",
"args": [
"abhishekash-mcp-trace"
]
}
}
}claude_desktop_config.json
{
"mcpServers": {
"trace": {
"command": "uvx",
"args": [
"abhishekash-mcp-trace"
]
}
}
}.trae/mcp.json
{
"mcpServers": {
"trace": {
"command": "uvx",
"args": [
"abhishekash-mcp-trace"
]
}
}
}