prompt

Free

SRE (Site Reliability Engineer) Agent prompt

FreeFree tier

Type

Open Source

Visit Website GitHub (40 stars)

About prompt

A specialized system prompt for an AI agent designed to act as a Site Reliability Engineer (SRE). It defines a data-driven personality focused on reliability as a feature with measurable budgets. The prompt includes detailed instructions for defining SLOs based on user experience, building observability across logs/metrics/traces, automating toil, conducting chaos engineering, and capacity planning. It also enforces rules like blameless culture, progressive rollouts, and using error budgets to drive decisions. The prompt provides concrete examples of SLO definitions, observability stacks (golden signals), and incident response integrations, making it suitable for deploying AI agents in production reliability roles.

Key Features

Defines SLOs and error budgets with measurable targets

Builds observability using logs, metrics, and traces (golden signals)

Automates toil to reduce repetitive operational work

Proactive chaos engineering to find weaknesses before users do

Data-driven capacity planning based on historical usage

Blameless post-incident reviews focusing on systemic fixes

Progressive rollouts with canary → percentage → full deployment

Incident response integration with SLO-based severity and automated runbooks

Pros & Cons

Pros

Data-driven and pragmatic about risk and trade-offs
Emphasizes blameless culture and systemic fixes
Provides concrete SLO examples and observability framework
Automation-obsessed to reduce toil and improve efficiency
Clear guidance on progressive rollouts and error budget consumption

Cons

Requires an AI model to execute; not a standalone tool
May be overly verbose for simple reliability scenarios
Assumes prior knowledge of SRE concepts (SLOs, burn rate)
Limited to the specific prompt structure; customization may be needed

Best For

Setting up and monitoring SLOs for microservices and APIsAutomating incident response and runbooks for known failure modesCapacity planning and right-sizing resources based on usage dataTraining AI assistants to follow SRE best practicesStandardizing reliability workflows across engineering teams

FAQ

What is the primary mission of this SRE agent?

To build and maintain reliable production systems through engineering, not heroics, focusing on SLOs, error budgets, observability, toil reduction, chaos engineering, and capacity planning.

What golden signals does the prompt recommend?

Latency (duration of requests), Traffic (requests per second), Errors (error rate by type), and Saturation (CPU, memory, queue depth, connection pool usage).

Does the prompt support incident response?

Yes, it includes severity based on SLO impact, automated runbooks for known failure modes, post-incident reviews focused on systemic fixes, and tracking MTTR.

What communication style does the agent use?

It leads with data (e.g., 'Error budget is 43% consumed'), frames reliability as investment, uses risk language, and is direct about trade-offs.