Mcp Victoriatraces
Give your agent authenticated access to VictoriaTraces so you can explore spans and trace APIs while fixing slow requests in production.
Overview
mcp-victoriatraces is an Operate-phase MCP server that links coding agents to VictoriaTraces trace APIs and documentation over authenticated stdio.
What is this MCP server?
- OCI package mcp-victoriatraces 1.5.0 with stdio transport to VT_INSTANCE_ENTRYPOINT
- Root URL for vtlingle or vtselect-style VictoriaTraces deployments (e.g. http://localhost:10428)
- Required VT_INSTANCE_BEARER_TOKEN for API authentication
- Optional MCP_DISABLED_TOOLS comma list to trim agent-exposed tools
- VictoriaTraces API plus documentation integration from VictoriaMetrics-Community
- Registry version 1.5.0
- OCI identifier ghcr.io/victoriametrics-community/mcp-victoriatraces:1.5.0
- Default local example entrypoint port 10428 documented in env description
Community signal: 9 GitHub stars.
What problem does it solve?
Chasing down slow endpoints forces you out of the agent into a separate trace UI because chat cannot query your VictoriaTraces backend.
Who is it for?
Builders with VictoriaTraces already ingesting OTLP or compatible traces who want MCP-driven trace investigation next to code changes.
Skip if: Projects with only metrics or logs, or teams standardized on a non-Victoria trace store without VictoriaTraces.
What do I get? / Deliverables
Registering the MCP lets your agent query and interpret VictoriaTraces data in-context while you patch the service causing the regression.
- Live MCP connection to VictoriaTraces
- Agent tools scoped to trace API and documentation
- Optional reduced tool surface via MCP_DISABLED_TOOLS
Recommended MCP Servers
Journey fit
Distributed tracing is an Operate concern once traffic is live—this MCP supports post-launch performance and reliability work. Monitoring covers trace backends; canonical placement is tracing observability alongside metrics and logs MCPs.
How it compares
Trace-store MCP bridge for VictoriaTraces, not metrics (mcp-victoriametrics) or log analytics (mcp-victorialogs).
Common Questions / FAQ
Who is mcp-victoriatraces for?
Developers operating VictoriaTraces who want AI agents to use the trace API and docs during production debugging.
When should I use mcp-victoriatraces?
Use it in Operate when investigating request latency, error propagation, or dependency failures that require distributed trace queries.
How do I add mcp-victoriatraces to my agent?
Configure the stdio MCP entry pointing at ghcr.io/victoriametrics-community/mcp-victoriatraces:1.5.0, set VT_INSTANCE_ENTRYPOINT and VT_INSTANCE_BEARER_TOKEN, optionally MCP_DISABLED_TOOLS, then reload MCP in your agent.