Agent Observability runs on your own machine — the client starts it, so there is no endpoint to ping. 53 installs a week from npm.
Agent observability: structured logging, cost tracking, and compliance audit trails
Nothing serious here today
Today is the operative word: we check Agent Observability every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
Connect this server
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add agent-observability -e 'AGENT_OBS_API_KEY=<AGENT_OBS_API_KEY>' -- npx -y agent-observability-mcp{
"mcpServers": {
"agent-observability": {
"command": "npx",
"args": [
"-y",
"agent-observability-mcp"
],
"env": {
"AGENT_OBS_API_KEY": "<AGENT_OBS_API_KEY>"
}
}
}
}[mcp_servers.agent-observability]
command = "npx"
args = ["-y", "agent-observability-mcp"]
env = { AGENT_OBS_API_KEY = "<AGENT_OBS_API_KEY>" }{
"mcpServers": {
"agent-observability": {
"command": "npx",
"args": [
"-y",
"agent-observability-mcp"
],
"env": {
"AGENT_OBS_API_KEY": "<AGENT_OBS_API_KEY>"
}
}
}
}{
"servers": {
"agent-observability": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"agent-observability-mcp"
],
"env": {
"AGENT_OBS_API_KEY": "<AGENT_OBS_API_KEY>"
}
}
}
}Put your own value in place of
<AGENT_OBS_API_KEY> (API key for the Agent Observability service (get one at POST /v1/register)).
The author’s registry entry marks it as required.
Optional settings from the registry entry, left out of the block above:
AGENT_OBS_API_URL (URL of the Agent Observability API (defaults to https://api-production-0c55.up.railway.app)).
Alternatives to Agent Observability
same job, measured the same wayStructured execution trace and span logging for AI agents
Structured observability for AI agents. Trace steps, decisions, and errors. MCP server + CLI viewer.
X.509 identity, risk scoring, and audit logging for AI agents. MiCA + EU AI Act compliant.
X.509 identity, risk scoring, and audit logging for AI agents. MiCA + EU AI Act compliant.
Observability, tracking, and root-cause for Security Wave scans and findings.
Query your AI agents' traces, errors, cost and alerts in Rius, the agent observability platform.
Structured Output MCP Traced Agent
Traced Structured Output Agent (Groq)
Agent Observability — questions
Answers built from our own checks of this server.