Pyxis

AI-powered infrastructure observability — log ingestion, distributed tracing, anomaly detection, and root cause analysis via Claude.

What it does

• Detects anomalies in logs and traces automatically
• Opens incidents and runs AI root cause analysis (RCA) using Claude
• APM / Tracing — tracks latency, p99, error rates per service; fires incidents on spikes
• Deployment correlation — links incidents to recent deploys
• Runbook generator — auto-writes a runbook when an incident resolves
• Log playground — paste any logs, get instant AI analysis
• Multi-tenant — one backend serves multiple teams

Quick install (any Linux)

curl -fsSL https://raw.githubusercontent.com/fropa/pyxis/main/install.sh | bash

The script will:

1. Install Docker if not present (Ubuntu, Debian, RHEL, Fedora, Arch)
2. Clone this repo
3. Ask for your Anthropic API key
4. Start the full stack
5. Create a default tenant and print your API key

Manual install

Requirements: Docker, Docker Compose

git clone https://github.com/fropa/pyxis.git
cd pyxis
cp backend/.env.example backend/.env
# Edit backend/.env and set ANTHROPIC_API_KEY
docker compose up -d --build

Then create a tenant to get your API key:

curl -X POST http://localhost:8000/api/v1/tenants/ \
  -H "Content-Type: application/json" \
  -d '{"name":"default"}'

Open http://localhost:5173 and paste the api_key from the response into Settings.

Services

Service	URL
Dashboard	http://localhost:5173
API	http://localhost:8000
API docs	http://localhost:8000/docs
PostgreSQL	localhost:5432
Redis	localhost:6379

Sending data

Logs (from any host, script, or CI pipeline):

curl -X POST http://localhost:8000/api/v1/ingest/ \
  -H "X-API-Key: YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "events": [{
      "source": "app_log",
      "level":  "error",
      "node_name": "web-01",
      "raw": "FATAL: database connection timeout after 30s"
    }]
  }'

Traces / APM (from any service):

curl -X POST http://localhost:8000/api/v1/traces/ \
  -H "X-API-Key: YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "spans": [{
      "trace_id":    "abc123",
      "span_id":     "def456",
      "service":     "api-gateway",
      "operation":   "GET /api/users",
      "duration_ms": 245,
      "status":      "ok",
      "status_code": 200
    }]
  }'

Deploy events (trigger correlation in RCA):

curl -X POST http://localhost:8000/api/v1/deploy-events/ \
  -H "X-API-Key: YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"service":"api-gateway","version":"v1.4.2","deployed_by":"ci"}'

Install agent on a Linux host

curl -fsSL http://YOUR_SERVER:8000/install | bash -s -- YOUR_KEY

The agent continuously ships logs and heartbeats from the host.

Stack

Layer	Technology
Backend	Python, FastAPI, SQLAlchemy, ARQ
AI	Claude (Anthropic API)
Database	PostgreSQL + pgvector
Cache	Redis
Frontend	React, Vite, Tailwind, Recharts
Deploy	Docker Compose / Helm (Kubernetes)

Environment variables

Variable	Description	Required
ANTHROPIC_API_KEY	Anthropic API key for Claude	Yes
DATABASE_URL	PostgreSQL connection string	Yes
REDIS_URL	Redis connection string	Yes
SECRET_KEY	Random secret for signing	Yes
DEBUG	Enable verbose SQL logging	No

License

MIT