Support

Open a support ticket and we'll get back to you within one business day (Standard) or four business hours (Enterprise). All traffic is one-on-one — no community forum, no shared Slack.

From the portal

The most-context-rich path. Go to /portal/support — the form captures the subject, the description, and a severity level. Replies arrive in-thread and we email you when one lands.

From the agent CLI

Useful when the agent itself is the thing that's broken. The ticket gets opened against your customer account, scoped to the specific agent, and (with --logs) attaches the last 200 lines of the agent log automatically:

bash

agentina report \
  --subject "<short summary>" \
  --message "<what happened, ideally with repro>" \
  --severity normal \
  --logs

--severity accepts low / normal / high / urgent. Default is normal.

Response times

Severity	Standard	Enterprise
`urgent`	4 business hours	1 hour, 24/7
`high`	1 business day	4 hours
`normal`	1 business day	1 business day
`low`	3 business days	3 business days

Reserve urgent for production-down. Liberal severities deplete our triage budget for everyone.

What to include

Symptoms. What you saw vs. what you expected.
Repro. Minimum sequence to reproduce. Even "I clicked X and it happened" is fine if that's really it.
Agent version + OS. agentina status prints both. The CLI flow attaches them automatically.
Last known good. If this used to work, when did it stop.
Log tail. The CLI flow attaches automatically. From the portal, paste the relevant lines into the description.

Security disclosures

For anything that looks like a vulnerability — key exposure, privilege escalation, supply-chain concern — email security@agentina.io directly instead of opening a public ticket. We acknowledge within one business day and disclose responsibly.

System status

Control plane liveness: https://cp.agentina.digcub-lab.com/v1/health. If this returns anything other than 200 ok and your agent is failing heartbeats, the problem is on our end and we're probably already paged.