SLO Error Budget
Performs deterministic SLO calculations: converts nines (e.g., 99.9%) to allowable downtime, computes error budgets and burn rates, generates humanized summaries without a metrics backend. SREs and DevOps engineers use it for reliability assessments in scripts, pipelines, or planning.
Overview
The SLO Error Budget MCP server executes precise, deterministic mathematics for Service Level Objectives (SLOs). It handles conversions from availability 'nines' to maximum downtime, error budget tracking, burn rate analysis, and human-readable status summaries. Designed for use without any metrics backend, it enables standalone computations directly in code or tools.
Key Capabilities
- nines_to_downtime: Converts SLO targets like 'three nines' (99.9%) to concrete downtime allowances (e.g., 8h 45m per month).
- error_budget: Calculates remaining error budget given target SLO, time period, and error count.
- burn_rate: Determines how quickly the error budget is being consumed based on recent errors.
- humanized_summary: Outputs plain-language explanations of SLO status, budgets, and risks.
These functions accept parameters like target percentage, period (monthly/quarterly), and error counts, returning structured results.
Use Cases
- Incident Analysis: After an outage, run nines_to_downtime(99.95, 'monthly') to quantify impact against SLO and share with teams.
- Deployment Gating: In CI/CD, query burn_rate(errors_last_7d) to block releases if burn exceeds 1.0x.
- Capacity Planning: Use error_budget(99.9, 'quarterly', projected_errors) to predict scaling needs.
- Reporting: Generate humanized_summary for dashboards or Slack bots to alert on budget exhaustion.
Who This Is For
SREs, DevOps engineers, and platform reliability teams managing production services. Useful for teams without full observability stacks or needing quick math in scripts, notebooks, or automation.