Site Reliability Engineering (SRE) Agent
Owns SLO/error-budget tracking, on-call runbook maintenance, and capacity planning for production services; performs autonomous investigation (root-cause correlation across logs/metrics/traces) on a live incident ahead of Incident Response Agent's own triage step, and recommends — but per established practice does not unilaterally execute — high-impact remediations.
How it runs
- What it reads
- SLO definitions, real-time metrics/logs/traces from Monitoring & Observability Agent, on-call runbook history, capacity/growth projections.
- What it produces
- error-budget status, autonomous root-cause investigation report, capacity forecast, runbook updates, remediation recommendation (human-approved for high-impact changes).
- When it runs
- continuous SLO monitoring, immediate investigation on any Monitoring & Observability Agent alert, scheduled capacity review monthly.
The same guardrails as every other agent
Site Reliability Engineering (SRE) Agent is governed exactly like the rest of the fleet. Its rule file is versioned, never edited in place. Its tool list is a declaration, not a grant — it can only call a tool a gateway policy has explicitly allowed in that project. Every tool call it makes is policy-checked and written to the audit log before it leaves the platform, and a project's cost cap is checked before a run starts, not reconciled after it.
More in DevOps & Release
- F1CI/CD & Infrastructure AgentMaintains pipelines and infrastructure-as-code (Terraform), manages environments.Read the product page
- F2Deployment/Release AgentExecutes releases, manages rollout strategy (blue-green/canary), and rollback procedures.Read the product page
- F3Monitoring & Observability AgentTracks logs, metrics, traces (Prometheus/Grafana-equivalent) and alerts on anomalies.Read the product page
- F4Incident Response AgentTriages production incidents, correlates logs, proposes/executes safe mitigations, pages humans when needed.Read the product page
- F5Backup & Disaster Recovery AgentVerifies backups actually restore, runs scheduled DR drills, and tracks RPO/RTO compliance against SLAs.Read the product page
- F6Chaos & Resilience Engineering AgentRuns scheduled fault-injection exercises ("GameDays") against staging/production-like environments to surface weak points before they cause real incidents — distinct from Backup & DR (restore verification of known-good backups) and Performance Testing (load only, no fault injection).Read the product page
- 192Docker Image Optimization Agentanalyzes and optimizes container images for size, build-cache efficiency, and attack surface; multi-stage build restructuring, minimal base image selection, layer-order optimization; distinct from CI/CD & Infrastructure Agent, which builds and maintains the pipeline that produces these images but doesn't analyze or optimize their own content, and from FinOps/Cost Optimization Agent, which tracks cloud billing spend, not the images themselves.Read the product page
- F8AIOps / Alert Correlation AgentIngests alert/event streams from Monitoring & Observability Agent and connected monitoring tools, correlates and deduplicates them into a smaller set of incidents using multi-signal matching (time window, topology/ownership key, text similarity), ranks probable root cause by evidence strength for any active incident, and recommends — but per established practice does not unilaterally execute — high-impact remediations; suppresses a root-cause candidate rather than guessing once its own confidence falls below a published threshold.Read the product page