Postmortems & reports
Postmortems drafted automatically that you can edit and share, plus weekly and monthly reports summarizing infrastructure health and trends.
- Editable Markdown drafts
- Weekly digest email
- Export as PDF
Product overview
The whole PrimesFlow platform, laid out.
Monitoring, AI-assisted diagnostics, alerting, incident history, and postmortems, built for teams managing their own infrastructure. Install in a minute, start seeing useful signals shortly after.
How it works
One agent per host, quietly reporting out.
Each host runs the PrimesFlow agent as an unprivileged service. It opens outbound HTTPS only, no inbound ports, no reverse tunnels.
Agent collects
10s resolution
Stream out
outbound HTTPS
PrimesFlow analyses
baseline + anomaly
You get context
cause + next steps
Real-time host monitoring
Every host, telling you the truth in real time.
CPU, memory, disk, network, systemd services, and Docker containers, collected down to 10-second resolution on the Team plan.
- Low-overhead agent
- Automatic host discovery
- Per-service telemetry
prod-api-01 · CPU
Fleet health
AI-assisted diagnostics
A plain-language summary, before you even ask.
When something spikes or breaks, PrimesFlow puts together a plain-language summary of what happened, what it touched, and what likely caused it.
- Metric + log correlation
- Suggested next steps
- Assistive — verify before acting
Threshold + anomaly alerts
Alerts that stay quiet until it matters.
Set static thresholds where you want control, or rely on anomaly detection where you don't. Alerts are grouped so one bad host doesn't page you fifty times.
- Email, Slack, Discord, webhooks
- PagerDuty & Opsgenie
- Noise-reduced grouping
Incident history
A complete timeline, every time.
Every incident is logged with a complete timeline: what tripped, what recovered, which alerts fired, and who acknowledged them.
- Chronological event log
- Filter by host or service
- Recurring-pattern detection
WarningINC-2481
Disk write rate above baseline
prod-db-01 · /var/log
More of the platform
Reports and install, wrapped up.
One-command install
One install script sets up the agent as an unprivileged systemd service. It only opens outbound HTTPS — no inbound ports.
- Ubuntu, Debian, RHEL, Rocky, Alma
- x86_64 and ARM64
- Uninstall in one command
Built to work across a team
Workspaces keep hosts and dashboards separate. Team-plan workspaces add role-based access so operators, developers, and contractors only see what they need.
- Owner / Admin / Member / Viewer roles on Team
- Per-workspace host grouping and tags
- Shared incident timeline and postmortems
- SSO available on Team plans
Run it against your own servers
14-day free trial, no credit card. One command to uninstall if it's not a fit.