ReliaStats
Fit failure data and compute MTBF, availability, and system reliability metrics inside an agent workflow for hardware or uptime-sensitive products.
Overview
ReliaStats is an Operate-phase MCP server that fits reliability distributions and computes MTBF, MTTR, availability, and composed system metrics for agents.
What is this MCP server?
- Weibull and lognormal distribution fitting from failure or survival data
- MTBF, MTTR, and availability calculations for components and systems
- System composition modeling for serial/parallel reliability structures
- Hosted streamable-http MCP remote at reliastats.com/mcp/v1
- Version 1.0.0 ReliaStats server with published server.schema.json manifest
- Server version 1.0.0
- One streamable-http remote: https://reliastats.com/mcp/v1
- Manifest schema: 2025-12-11 server.schema.json
What problem does it solve?
You need defensible reliability numbers from failure data but do not want to leave the agent chat for specialized stats software or hand-rolled notebooks.
Who is it for?
Indie builders working on hardware, robotics, industrial SaaS, or any product where uptime and failure modes are part of the pitch or ops loop.
Skip if: Typical CRUD web apps or marketing sites where reliability statistics never appear in requirements or incident reviews.
What do I get? / Deliverables
Your agent can call ReliaStats remotes to return fitted parameters and standard reliability KPIs you can paste into runbooks, specs, or investor updates.
- Fitted Weibull or lognormal parameters from your inputs
- MTBF, MTTR, and availability figures for components or systems
- Agent-readable reliability summaries for docs or incident reports
Recommended MCP Servers
Journey fit
Reliability math belongs on the Operate shelf because it supports production health, failure analysis, and uptime planning after something is built or deployed. Monitoring is the canonical subphase for MTBF/MTTR, availability, and distribution fitting used to interpret real-world failure and repair signals.
How it compares
Remote reliability-statistics MCP, not a general monitoring dashboard or a Claude skill for code review.
Common Questions / FAQ
Who is com.reliastats/public for?
Builders and operators who need Weibull/lognormal fitting, MTBF/MTTR, and availability math inside Claude Code, Cursor, or similar MCP clients.
When should I use ReliaStats?
Use it during Operate when you are analyzing failures, modeling redundancy, or documenting SLA-related metrics with an agent.
How do I add ReliaStats to my agent?
Register the streamable-http remote https://reliastats.com/mcp/v1 in your MCP client config using the published server manifest (version 1.0.0).