Spoor
Read-only MCP server that ranks what changed across metrics, logs and traces
- Pricing
- Open source
- Platforms
- macOS, Linux, Windows
- Runs
- Locally
- Version
- 0.3.1
- Listing updated
- 2 days ago
About Spoor
Spoor is an open-source MCP server for developer tools. It runs locally on your machine and is started by your AI client.
Spoor does the comparison work an agent would otherwise do with raw queries. It looks at one time window against a baseline across request rate, error rate, latency, CPU, memory and consumer lag, plus new and spiking error logs and slower or failing spans. The result is a short ranked answer, typically a few hundred tokens, with the service where a failure started on top.
Callers that only suffer from a broken dependency are labelled as originating elsewhere (for example "originates in payment") and moved down. Errors that were already present are listed as steady rather than new, and a coverage section says when a backend returned no data, so "no change" is never confused with "can't see".
The README describes it as deterministic Go with no LLM inside, read-only, with emails, tokens and IPs scrubbed from output. Backends are VictoriaMetrics, VictoriaLogs, VictoriaTraces, Prometheus, Loki, Tempo and New Relic, chosen through environment variables. It ships as a static binary, a Go install or an OCI image under Apache-2.0.
Install
Run in a terminal
claude mcp add spoor --env SPOOR_METRICS_TOKEN='<SPOOR_METRICS_TOKEN>' --env SPOOR_LOGS_TOKEN='<SPOOR_LOGS_TOKEN>' --env SPOOR_TRACES_TOKEN='<SPOOR_TRACES_TOKEN>' --env NEW_RELIC_API_KEY='<NEW_RELIC_API_KEY>' -- docker run -i --rm -e SPOOR_METRICS_TOKEN -e SPOOR_LOGS_TOKEN -e SPOOR_TRACES_TOKEN -e NEW_RELIC_API_KEY ghcr.io/asinha07/spoor:0.3.1Add to claude_desktop_config.json
{
"mcpServers": {
"spoor": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"SPOOR_METRICS_TOKEN",
"-e",
"SPOOR_LOGS_TOKEN",
"-e",
"SPOOR_TRACES_TOKEN",
"-e",
"NEW_RELIC_API_KEY",
"ghcr.io/asinha07/spoor:0.3.1"
],
"env": {
"SPOOR_METRICS_TOKEN": "<SPOOR_METRICS_TOKEN>",
"SPOOR_LOGS_TOKEN": "<SPOOR_LOGS_TOKEN>",
"SPOOR_TRACES_TOKEN": "<SPOOR_TRACES_TOKEN>",
"NEW_RELIC_API_KEY": "<NEW_RELIC_API_KEY>"
}
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"spoor": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"SPOOR_METRICS_TOKEN",
"-e",
"SPOOR_LOGS_TOKEN",
"-e",
"SPOOR_TRACES_TOKEN",
"-e",
"NEW_RELIC_API_KEY",
"ghcr.io/asinha07/spoor:0.3.1"
],
"env": {
"SPOOR_METRICS_TOKEN": "<SPOOR_METRICS_TOKEN>",
"SPOOR_LOGS_TOKEN": "<SPOOR_LOGS_TOKEN>",
"SPOOR_TRACES_TOKEN": "<SPOOR_TRACES_TOKEN>",
"NEW_RELIC_API_KEY": "<NEW_RELIC_API_KEY>"
}
}
}
}Add to .vscode/mcp.json
{
"servers": {
"spoor": {
"type": "stdio",
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"SPOOR_METRICS_TOKEN",
"-e",
"SPOOR_LOGS_TOKEN",
"-e",
"SPOOR_TRACES_TOKEN",
"-e",
"NEW_RELIC_API_KEY",
"ghcr.io/asinha07/spoor:0.3.1"
],
"env": {
"SPOOR_METRICS_TOKEN": "<SPOOR_METRICS_TOKEN>",
"SPOOR_LOGS_TOKEN": "<SPOOR_LOGS_TOKEN>",
"SPOOR_TRACES_TOKEN": "<SPOOR_TRACES_TOKEN>",
"NEW_RELIC_API_KEY": "<NEW_RELIC_API_KEY>"
}
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"spoor": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"SPOOR_METRICS_TOKEN",
"-e",
"SPOOR_LOGS_TOKEN",
"-e",
"SPOOR_TRACES_TOKEN",
"-e",
"NEW_RELIC_API_KEY",
"ghcr.io/asinha07/spoor:0.3.1"
],
"env": {
"SPOOR_METRICS_TOKEN": "<SPOOR_METRICS_TOKEN>",
"SPOOR_LOGS_TOKEN": "<SPOOR_LOGS_TOKEN>",
"SPOOR_TRACES_TOKEN": "<SPOOR_TRACES_TOKEN>",
"NEW_RELIC_API_KEY": "<NEW_RELIC_API_KEY>"
}
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"spoor": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"SPOOR_METRICS_TOKEN",
"-e",
"SPOOR_LOGS_TOKEN",
"-e",
"SPOOR_TRACES_TOKEN",
"-e",
"NEW_RELIC_API_KEY",
"ghcr.io/asinha07/spoor:0.3.1"
],
"env": {
"SPOOR_METRICS_TOKEN": "<SPOOR_METRICS_TOKEN>",
"SPOOR_LOGS_TOKEN": "<SPOOR_LOGS_TOKEN>",
"SPOOR_TRACES_TOKEN": "<SPOOR_TRACES_TOKEN>",
"NEW_RELIC_API_KEY": "<NEW_RELIC_API_KEY>"
}
}
}
}Configuration
SPOOR_METRICS_TYPEBackend for metrics: victoriametrics (default), prometheus or newrelicSPOOR_LOGS_TYPEBackend for logs: victorialogs (default), loki or newrelicSPOOR_TRACES_TYPEBackend for traces: victoriatraces (default), tempo or newrelicSPOOR_METRICS_URLMetrics backend URL (VictoriaMetrics or Prometheus) as seen from the container, e.g. http://host.docker.internal:8428SPOOR_LOGS_URLLogs backend URL (VictoriaLogs or Loki) as seen from the container, e.g. http://host.docker.internal:9428SPOOR_TRACES_URLTraces backend URL (VictoriaTraces or Tempo) as seen from the container, e.g. http://host.docker.internal:10428SPOOR_METRICS_TOKENBearer token for the metrics backend, if it requires oneSPOOR_LOGS_TOKENBearer token for the logs backend, if it requires oneSPOOR_TRACES_TOKENBearer token for the traces backend, if it requires oneNEW_RELIC_ACCOUNT_IDNew Relic account ID, for signals of type newrelicNEW_RELIC_API_KEYNew Relic User API key (NRAK-...), for signals of type newrelic
What it does
- Cross-signal change ranking
- Compares metrics, logs and traces for the same window against the same baseline.
- Cause versus echo
- Uses trace trees to show where failures start and demotes callers that only suffer.
- Steady errors called out
- Lists errors that were already there as not new, so agents stop blaming them.
- Coverage reporting
- States when a backend failed or had no data instead of reporting no change.
- Token budget
- Every tool accepts budget_tokens, default 800, to bound the answer size.
- Mixed backends
- Switch metrics, logs and traces between Victoria, Prometheus, Loki, Tempo or New Relic.
Good for
- Find which service an incident started in after an alert fires
- Let an AI agent triage a failing deploy using a few tool calls
- Separate new error patterns from errors that were always present
- Check New Relic or Grafana-stack telemetry from an MCP client
Pricing
Open source
Apache-2.0; free to run yourself
Prices change. Check telespoor.com before you buy.
Spoor FAQ
Does Spoor use an LLM internally?
No. The README says it is deterministic Go, so the same input gives the same output.
Can it change anything in my systems?
No. It is read-only and only calls query endpoints.
Which backends does it support?
VictoriaMetrics, Prometheus and New Relic for metrics; VictoriaLogs, Loki and New Relic for logs; VictoriaTraces, Tempo and New Relic for traces.
How do I run it?
As a static binary for macOS, Linux and Windows, with go install, or as the ghcr.io/asinha07/spoor Docker image that speaks MCP over stdio.
What licence is it under?
Apache-2.0.
Is Spoor free?
Yes. Spoor is open source, so you can use and self-host it for free.
How do I add Spoor to Claude Code?
Run this command in your terminal: claude mcp add spoor --env SPOOR_METRICS_TOKEN='<SPOOR_METRICS_TOKEN>' --env SPOOR_LOGS_TOKEN='<SPOOR_LOGS_TOKEN>' --env SPOOR_TRACES_TOKEN='<SPOOR_TRACES_TOKEN>' --env NEW_RELIC_API_KEY='<NEW_RELIC_API_KEY>' -- docker run -i --rm -e SPOOR_METRICS_TOKEN -e SPOOR_LOGS_TOKEN -e SPOOR_TRACES_TOKEN -e NEW_RELIC_API_KEY ghcr.io/asinha07/spoor:0.3.1
What are the best alternatives to Spoor?
The closest alternatives to Spoor are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.
Alternatives to Spoor

REA
MCP serverCLI and MCP server that lets coding agents reverse engineer apps and binaries
Runs locally- Open source
- 464K GitHub stars
- Developer tools

Codebase Memory
MCP serverLocal MCP server that builds a code knowledge graph for AI coding agents
Runs locally- Open source
- 48.7K GitHub stars
- Knowledge and memory

GitHub
MCP serverOfficial MCP server connecting AI assistants to GitHub repos, issues and workflows
Not checked yet- Remote
- Open source
- 34.5K GitHub stars
- Developer tools

Serena MCP: the IDE for your agent
MCP serverMCP toolkit giving coding agents symbol-level code retrieval and editing
Runs locally- Open source
- 31.5K GitHub stars
- Developer tools

XERJ
MCP serverLocal search engine for AI agents with cited answers over your code and docs
Runs locally- Open source
- 19.2K GitHub stars
- Knowledge and memory

Agent-Native Analytics
MCP serverOpen-source analytics app where an AI agent builds charts and dashboards from your data
- Remote
- Open source
- 7.7K GitHub stars
- Analytics and data