AutoOps AI

One command to autonomously configure Splunk observability and investigate incidents for any codebase.

Install (no repo clone required)

# macOS / Linux (recommended)
pipx install autoops-ai
# or from source
curl -fsSL https://raw.githubusercontent.com/YOUR_ORG/autoops-ai/main/scripts/install.sh | bash

# Windows PowerShell
pip install autoops-ai
# or: .\scripts\install.ps1

First Run

autoops setup

Interactive wizard lets you choose:

• OpenAI
• Claude (Anthropic)
• OpenRouter
• Azure OpenAI
• Amazon Bedrock (Claude)

Credentials are saved securely in your OS keychain (macOS Keychain / Windows Credential Locker).

Configure Any Project

cd /path/to/your/app
autoops configure --repo .

AutoOps will:

1. Detect/start Docker Desktop (macOS + Windows)
2. Pull and start Splunk Enterprise (autoops-splunk)
3. Start OpenTelemetry Collector (autoops-otel-collector)
4. Scan your repo and discover architecture
5. Generate instrumentation
6. Create Splunk dashboards, saved searches, and alerts
7. Send a synthetic telemetry test event

Command Reference

autoops init --repo <path> is an alias for configure.

Splunk Defaults Created

Dashboards: Service Health Overview, Database Performance, Deployment Timeline, Incident Investigation, per-service health

Saved searches: error rate, latency, p95 latency, DB latency, deployments, no-data detection

Alerts: per-service error rate, checkout error rate, 5xx spike, ingestion stopped

Sample SPL to customize:

index=main sourcetype=autoops | stats count by service, path, level
index=main sourcetype=autoops path="/checkout" | timechart avg(duration_ms) p95(duration_ms)

Prerequisites

• Python 3.11+
• Docker Desktop (AutoOps detects and opens installer if missing)
• LLM API key (chosen during autoops setup)

Demo App

autoops demo start
cd demo-app && uvicorn main:app --port 8080
autoops demo bug-on
autoops demo traffic --requests 30
autoops investigate --alert checkout_error_rate

License

MIT