Aurora

Open-source AI-agent incident management and root cause analysis: it investigates your stack automatically when an alert fires and delivers a structured RCA before you open your laptop.

Source repo
Arvo-AI/aurora
Stars
★ 465
Last updated
4d ago
License
Apache-2.0
Primary language
Python

At a glance

How it runs
Self-hosted serviceWeb appMCP server
Works with
Universal · cross-platformOpenAI API · Claude API
Cost
Free software; you pay for model usage
Setup effort
High · needs real infrastructure
You'll need
DockerKubernetes (for sandboxed agent execution and Helm deployment)HashiCorp VaultPostgreSQLRedisMemgraphLLM API key (OpenRouter, OpenAI, etc.)Shell / CLINetwork accessLocal filesystemMCP Server
Typical use
An SRE team paged by PagerDuty or Datadog at 3 AM wants the alert auto-triaged and investigated in the background instead of waking an engineer for a 30-60 minute manual investigation.
Not a fit if
  • Teams that cannot self-host Docker/Kubernetes infrastructure
  • Organizations unwilling to grant LLM agents access to cloud and cluster credentials
  • Users who want a vendor-hosted SaaS that works out of the box
Website
aurorasre.ai ↗ set on GitHub
Maintenance
Commits in the last 30 dayslast commit 2026-11-16latest release v1.5.25 (2026-11-16)+0.6 ★/day over 7 days

What does this agent do, and when should you use it?

Aurora is an open-source, AI-agent-powered incident management platform from Arvo-AI, built for SRE and DevOps teams. When an alert arrives, LangGraph agents dynamically pick from 30+ tools and run kubectl, aws, az, and gcloud inside sandboxed Kubernetes pods, querying logs, checking deployments, and correlating data across AWS, Azure, GCP, and Kubernetes to produce a structured root cause analysis. The system consists of a Python/Flask backend with Celery workers, a Next.js frontend, Memgraph for the infrastructure knowledge graph, Vault for secrets, and PostgreSQL/Redis storage, deployed via Docker Compose or Helm entirely on your own infrastructure with zero telemetry. It also auto-generates postmortems exportable to Confluence, Notion, or SharePoint, traces blast radius across services through a dependency graph, and can generate pull requests with remediation fixes.

Aurora ingests alerts from PagerDuty, Datadog, Grafana, New Relic, OpsGenie, incident.io, and CloudWatch alarm webhooks; every alert auto-triggers a background investigation. LangGraph agents select from 30+ tools and run kubectl, aws, az, and gcloud in isolated Kubernetes pods (with NetworkPolicy), query logs, check deployments, and correlate data across services and providers; investigations draw on Knowledge Base RAG and the Memgraph-backed infrastructure graph for blast radius analysis. On completion, Aurora produces an RCA report with timeline, root cause, impact assessment, and remediation steps, can auto-generate a postmortem and export it to Confluence/Notion/SharePoint, generate a fix pull request, and run automated Actions (open PRs, notify Slack). Persistent Artifacts documents are continuously updated as the investigation progresses; the platform also exposes an MCP Server (for Cursor, Claude Desktop, Windsurf), Terraform/IaC analysis, SigmaHQ command guardrails (37 threat detection signatures), and a NeMo input rail.

  1. An SRE team paged by PagerDuty or Datadog at 3 AM wants the alert auto-triaged and investigated in the background instead of waking an engineer for a 30-60 minute manual investigation.
  2. Multi-cloud teams need one agent to investigate incidents spanning AWS, Azure, GCP, Kubernetes, or even Fly.io and trace blast radius across services.
  3. Engineering leads want auto-generated postmortems with timeline and impact assessment exported straight to Confluence, Notion, or SharePoint.
  4. Platform teams want to capture investigation reasoning so knowledge stops living in individual engineers' heads.
  5. Security-sensitive organizations require incident data to never leave their infrastructure, with Ollama support for fully air-gapped operation.
  6. MCP users want to invoke Aurora's investigation capabilities directly from Cursor or Claude Desktop.

Key setup facts

Pulled from the install and usage notes below. The project README has the current values.

Ports
3000
Environment variables
VAULT_TOKEN
Config files
.env my-values.yaml
Install notes cover
Docker

How do you install or deploy this agent?

Local evaluation in under 5 minutes (Docker required):

bash

git clone https://github.com/arvo-ai/aurora.git && cd aurora
make init                # Generate secure secrets
nano .env                # Add your LLM API key (OpenRouter, OpenAI, etc.)
make prod-prebuilt       # Pull prebuilt images and start

Open http://localhost:3000; the first registered user becomes admin. Vault setup is required after first start:

bash

# Get the auto-generated root token
docker logs vault-init 2>&1 | grep "Root Token:"
# Add it to .env
echo "VAULT_TOKEN=hvs.your-token-here" >> .env
# Restart to connect services to Vault
make down && make prod-prebuilt

Other options: pin a version with make prod-prebuilt VERSION=v1.2.3; build from source with make prod-local; production Kubernetes via Helm:

bash

helm repo add aurora https://raw.githubusercontent.com/Arvo-AI/aurora/gh-pages
helm repo update
helm show values aurora/aurora-oss > my-values.yaml

# Edit my-values.yaml, then:

helm install aurora-oss aurora/aurora-oss -n aurora --create-namespace -f my-values.yaml

Also available via OCI: oci://ghcr.io/arvo-ai/charts/aurora-oss. An air-tight bundle covers air-gapped / restricted networks.

How do you use this agent?

After deployment, open http://localhost:3000; the first registered user is admin. The only required external configuration is an LLM API key in .env (OpenRouter, OpenAI, Anthropic, Gemini, Vertex AI, AWS Bedrock, or self-hosted Ollama); all cloud connectors are optional and Aurora runs without any cloud provider accounts. Day-to-day flow: connect PagerDuty, Datadog, Grafana, New Relic, OpsGenie, incident.io, or CloudWatch alarm webhooks; alerts automatically trigger background investigations where agents run kubectl/aws/az/gcloud in sandboxed pods and produce an RCA. Admins can explore the infrastructure knowledge graph, export investigations as postmortems, and configure Actions to open fix PRs or notify Slack automatically when investigations complete.

What are this agent's strengths and limitations?

Pros
  • Alerts auto-trigger full-stack investigations: agents run kubectl/aws/az/gcloud in sandboxed Kubernetes pods with NetworkPolicy, off the control plane, backed by 37 SigmaHQ threat signatures and a NeMo prompt-injection rail.
  • End-to-end closure: RCA reports, auto-generated postmortems (export to Confluence/Notion/SharePoint), fix PR generation, and post-RCA Actions workflows.
  • Model- and cloud-agnostic: OpenAI, Anthropic, Gemini, Vertex AI, AWS Bedrock, OpenRouter, and Ollama (air-gapped); connectors for AWS/Azure/GCP/OVH/Scaleway/Cloudflare.
  • 100% self-hosted with zero telemetry, secrets encrypted at rest in Vault or AWS Secrets Manager, Apache 2.0 with no per-seat or per-incident pricing.
Limitations
  • High deployment overhead: Docker Compose or Kubernetes/Helm plus Vault, PostgreSQL, Redis, and Memgraph, with a manual Vault root-token configuration step after first start.
  • Ongoing LLM costs: agents call many tools and models in parallel per alert; you supply paid API keys and the source gives no usage estimates.
  • Security sensitivity: agents need credentials to execute cloud CLIs and kubectl, so routing an LLM into your cloud credential path is itself a new attack surface to evaluate.
  • As a young project, public production case studies or third-party audits beyond README/docs/Discord are not evidenced.

How does this agent compare with similar options?

Key facts side by side with the most closely related agents.

Agent Source review Form / cost Stars Updated Language Full support on
Aurora This agent 54 · Major gaps Self-hosted serviceFree + model costs ★ 465 4d ago Python OpenAI API · Claude API
HolmesGPT SRE Agent 62 · Some gaps CLIFree + model costs ★ 3.9k 3d ago Python OpenAI API
Ongrid 70 · Some gaps Self-hosted serviceFree + model costs ★ 1.3k 2d ago Go OpenAI API · Claude API
AURA SRE Agent Platform 72 · Some gaps CLIFree + model costs ★ 414 21d ago Rust OpenAI API

How does FollowAgents rate this agent?

FollowAgents source review · FARS-2.1
Major gaps
54/ 100 5-point scale 2.7 / 5
This review covers the source as of Nov 08, 2026. The project has shipped 12 stable release(s) since, which have not been re-reviewed. The score and verdict describe the reviewed version only. Recent releases ↓
Trust 12/29
Reliability 6/14
Adaptability 14/18
Convention 11/18
Effectiveness 7/13
Verifiability 4/8
Why each dimension lost points
Trust12 / 29 · 2.1/5

The README says commands run in isolated Kubernetes pods and describes RBAC, Vault, command policies, and input safeguards. AGENTS.md gives detailed fail-closed rules for Ask/RCA read-only command classification, supporting partial least-privilege evidence. However, the sampled kubectl-agent.py executes commands received over its authenticated WebSocket without showing agent-side authorization or confirmation. The Bitbucket and Cloudflare skills include writes such as editing or merging code, triggering pipelines, changing DNS, and purging caches, while confirmation and recovery procedures are limited in the supplied material. The README describes self-hosted data, direct LLM-provider calls, no telemetry, and Vault or AWS Secrets Manager storage for secrets, but the sample does not independently verify those claims. No dependency security audit evidence is supplied. The repository and security contact are identifiable, but publisher identity remains unknown as specified.

Reliability6 / 14 · 2.1/5

AGENTS.md and the README broadly agree on architecture and security features, but the sampled agent source shows only connection, reconnection, and kubectl execution paths, so it cannot substantiate the product-wide claims. The code logs connection and send failures and returns structured command errors, including timeouts and exit codes. Dependency availability is supported only in part by reconnection behavior and deployment notes; the sample does not establish dependency pinning or availability guarantees.

Adaptability14 / 18 · 3.9/5

The README clearly targets SRE incident response and describes cloud, Kubernetes, monitoring, alerting, and CI/CD scenarios. It offers local, Helm, and air-gapped deployment paths and several LLM choices. Skill files name tools, parameters, and connection conditions, while the command-classification guidance distinguishes Ask/RCA from Agent mode. The sample does not fully establish when autonomous investigations, automatic triggers, or code fixes are enabled or what controls users have over them. Multiple deployment options support different environments, but some capabilities depend on external services and configuration.

Convention11 / 18 · 3.1/5

The README is organized around features, quick start, deployment, integrations, security, privacy, and architecture; AGENTS.md and CLAUDE.md provide developer and operations guidance. The quick start covers initialization, secret configuration, startup, and a Vault follow-up, and the skills include practical examples. Deductions reflect that the guidance files are mainly contributor and operations material rather than a complete product specification, and the sample does not cover the whole documentation set. Naming stability cannot be established across a few sampled skills and tools. Some limitations are documented, including the LLM requirement, optional connectors, and CloudBees connection prerequisites, but coverage is partial. The full Apache 2.0 license is present, warranting full marks. The README points to releases and a changelog, but the changelog contents are not in the sample. Maintenance channels and vulnerability reporting are specified, but there is no clear long-term maintenance commitment, and publisher identity is unknown.

Effectiveness7 / 13 · 2.7/5

The product connects alert intake, investigation, root cause analysis, knowledge capture, and postmortems. The README describes structured RCA, suggested fixes, and integrations, while the skills provide concrete investigation steps. The command agent returns success state, output, error, and return code; the sample does not show the RCA output format or how it separates evidence from inference, so output usability is only partially supported. Automated investigations and write actions could save engineering time, but the multi-service deployment, Vault, database, and external LLM costs are not quantified in the supplied material.

Verifiability4 / 8 · 2.5/5

Some README claims about quick start, security, and architecture have corresponding details in AGENTS.md, skill files, and the agent source; documentation also identifies implementation and test paths. However, the referenced tests are not supplied, the build validation workflow is marked temporarily paused, and the sample lacks enough source to cross-check key security and capability claims. The README points to public benchmark material and says losses are included, but those results are not part of this packet. Claims are generally separated from operational examples, and some limitations are stated; several privacy and security claims still lack independent corroboration in the supplied evidence.

Risks and how to mitigate them
  • Before connecting cloud accounts, clusters, or code hosts, review connector permissions and identify which investigation or remediation actions write to external systems; the sample does not show the kubectl agent's complete authorization path.
  • Treat the README's self-hosting, no-telemetry, and security-control claims as not independently verified by this static sample; the build validation workflow is marked temporarily paused.
Evidence confidence: Low Reviewed Nov 08, 2026 Reviewed revision d2c193a9ed73 New commits since this review; the score may not cover them
See the full review method →
Maintain this project? Add the FARS badge to your README

The badge shows the current review score and verdict and links back to this page. It refreshes about once an hour after a re-review.

FARS

[![FARS review](https://followagents.com/badge/aurora.svg)](https://followagents.com/en/agents/aurora)
<a href="https://followagents.com/en/agents/aurora"><img src="https://followagents.com/badge/aurora.svg" alt="FARS review"></a>

Recent releases

Latest entries from the project's GitHub Releases. Notes are quoted from upstream as written, not rewritten or tested by FollowAgents.

  1. v1.5.25
    Release notes excerpt
    Automated release v1.5.25
    View on GitHub ↗
  2. aurora-oss-1.5.25
    Release notes excerpt
    Aurora – AI-powered cloud operations platform (on-prem deployment)
    View on GitHub ↗
  3. v1.5.24
    Release notes excerpt
    Automated release v1.5.24
    View on GitHub ↗

FAQ

Do I need cloud provider accounts to use it?
No. The README states the LLM API key is the only external requirement; all cloud/Kubernetes connectors are optional. It deploys and runs without connectors, though investigation scope will be limited.
Is it safe to let agents run commands?
Commands run in isolated Kubernetes pods with NetworkPolicy, not on your control plane; SigmaHQ provides 37 threat-detection signatures on agent command execution, a NeMo input rail detects prompt injection, and org-level command policies let you restrict behavior.
Does my data go to Arvo AI?
No. Aurora is 100% self-hosted with zero telemetry; LLM calls go directly from your infrastructure to your chosen provider, and Ollama enables fully air-gapped operation.
Which LLMs are supported, and what does it cost?
OpenAI, Anthropic, Gemini, Vertex AI, AWS Bedrock, OpenRouter, and Ollama are supported. The software is free under Apache 2.0, but you pay your chosen model provider for API usage.
How do I deploy to production?
Use the official Helm chart (GKE/EKS/AKS, also installable from the OCI registry) or the air-tight bundle for air-gapped networks; local evaluation uses make prod-prebuilt.
Is Aurora open source?
Yes. The source code is on GitHub (Arvo-AI/aurora) under the Apache-2.0 licence.
Is Aurora still maintained?
Yes. The latest commit on GitHub is from 2026-11-16, within the last 30 days. The latest release is v1.5.25 (2026-11-16).
Who makes Aurora?
Aurora is published on GitHub by the account Arvo-AI. FollowAgents has not verified which person or company is behind that account.
What is Aurora's official website?
The project's GitHub repository lists https://www.aurorasre.ai as its website. That link is set by the repository owner; the source code itself is on GitHub (Arvo-AI/aurora).
Was this page helpful?
View on GitHub ↗ Install ↓

Related concepts