Release state
What's shipped, what's beta, what's planned
We don't want you to guess what you're buying. Every Vigilhold feature is labeled with its actual shipping state. If it isn't marked LIVE, it isn't covered by our support response targets.
Current release: v5.15.0 · Updated 2026-11-04 · View changelog →- LIVE
- Shipped, stable, covered by support response targets
- BETA
- Functional. Interface or behavior may change before stable.
- EXP
- Early access. Breaking changes possible. Not covered by support response targets.
- ROADMAP
- Planned but not yet shipped. No ETA unless stated.
Detection & Response
19 LIVE 2 ROADMAP LIVEAgent offline warnings
Servers that stop reporting are marked with how long they have been silent, counted in the fleet header, and listed under Needs attention.
All plans LIVEAgent installation (one-line command)
A single command installs the Vigilhold agent. It connects out to Vigilhold, so there are no inbound ports to open.
All plans LIVEProcess, port, and network monitoring
CPU, memory, open ports, running processes, and network connections, reported continuously by the agent.
All plans LIVESSH brute-force detection
Detects SSH brute force on the host and bans the source.
All plans LIVEPort-scanner bans
Bans addresses that probe the server for open ports, enforced in the kernel. Trusted addresses are never banned and open sessions are never cut.
All plans LIVEAutonomous IP ban
Blocks attacker addresses on the host when a threat is confirmed. Reversible, and logged with the reason.
All plans LIVEAutonomous process kill
Stops a process that matches a threat without waiting for a person, at Autopilot and above. Every kill is logged; a kill cannot be undone.
Business+ LIVELockdown mode
Restricts inbound and outbound network access on a server while investigation continues.
All plans LIVEAlert rules (plain English)
Write CPU, memory, and disk alert rules in plain English. They are turned into structured conditions, and rules that contradict each other are flagged.
All plans LIVEPlaybooks
Choose what happens for each kind of threat: ban the source, lock the server down, or alert only.
All plans LIVEResponse rules
Rules that respond in steps: ban, stop a process, quarantine, or lock down. Notify-only rules are on every plan.
Business+ LIVEAudit log (tamper-evident)
Actions are written to a hash chain, so an edited or deleted entry shows up when the chain is verified.
All plans LIVEApproval queue
Actions that need a person surface for approval before they run.
All plans LIVEUndo for autonomous actions
Bans, network isolation, lockdown, and file quarantine can be undone within the undo window. Process kills cannot.
All plans LIVEDDoS detection
Detects SYN floods, connection floods, UDP amplification, and HTTP floods. Single-source connection floods are banned automatically.
All plans LIVEWorm detection
Detects worm-like behavior on a server, such as spreading to other hosts or copying itself.
All plans LIVECanary files (tripwires)
Decoy files in system directories on each server that nothing legitimate touches. Changing one raises an alert, and so does opening one where inotify-tools is installed (the installer adds it).
Business+ LIVEDocker container controls
Start, stop, restart, pause, or unpause containers from the dashboard. Admin only, and audited.
All plans ROADMAPKubernetes workload monitoring
Not live yet. Planned: flags risky workloads (privileged containers, host mounts, host networking) and failing ones (crash loops, out-of-memory kills).
All plans ROADMAPKubernetes namespace isolation
Not live yet. Planned: when a container escape is detected, the namespace is isolated with a deny-all network policy.
All plans LIVEAgent config resync on reconnect
After a disconnect, agents automatically fetch the current ban list and process allowlist from Vigilhold.
All plansFleet Intelligence
10 LIVE 1 ROADMAP LIVEMulti-stage attack tracking
Links events on a host into the stages of an attack (MITRE ATT&CK), so a slow intrusion reads as one incident rather than scattered alerts.
All plans LIVEFleet-wide warnings
When one server catches an attacker, the rest of your own fleet is warned. On by default; nothing leaves your organization.
All plans LIVEThreat sharing between organizations
Off unless you opt in (Fleet → Threat intel). Organizations that opt in share attacker addresses and attack types with each other, never hostnames or anything else about their servers.
All plans LIVEOffline defense
When Vigilhold is unreachable, agents keep enforcing their offline defense policy and report what happened once they reconnect.
All plans LIVEProactive threat sweep
Vigilhold regularly checks each server for signs of compromise the agent has not raised on its own, and opens an investigation when it finds one.
Enterprise LIVEAnomaly detection (behavioral baseline)
Learns each server's normal CPU, memory, and disk behavior and alerts on unusual spikes.
Business+ LIVESmart Tune (suggested alert rules)
Suggests alert rules from each server's learned baseline. Pick one to load it into the rule editor.
Business+ LIVEThreat map
Map of the addresses you have banned and where they come from. Location lookup is opt-in.
All plans LIVEFleet threat heatmap
Every server as a tile colored by its early-warning score, with links where lateral movement connects hosts.
All plans LIVEFleet Timeline (MITRE ATT&CK stages)
Alerts from every server on one timeline on the Hosts page, each labeled with its MITRE ATT&CK stage.
All plans ROADMAPNL Query (natural language fleet search)
Not available right now. Planned: ask about your fleet in plain English and get a structured answer. Scoped to one server or the whole fleet.
All plansCompliance & Evidence
3 LIVE LIVECompliance reports (CIS, SOC 2, PCI DSS)
Point-in-time evidence against CIS, SOC 2, and PCI DSS, with pass/fail per control.
Business+ LIVECompliance remediation
Finds drift in supported controls on a schedule and makes the fix once you approve it.
Business+ LIVEContinuous audit export
Export the audit log as an encrypted bundle, or send it to S3-compatible storage. CSV and JSON export is on every plan. The hash chain can be verified independently.
EnterpriseIdentity & Access
7 LIVE LIVESSH key management
Compare each server's authorized_keys against a known baseline, and add or remove keys per server.
All plans LIVECredential vault (AES-256)
AES-256-GCM encrypted store with a separate key for each organization. Reveals are admin only, rate limited, and audited.
All plans LIVEEmergency access tokens
Single-use, time-limited links for a locked-out admin, giving a read-only session. Issue and use are audited, and the token is stripped from the address bar on load.
Business+ LIVETwo-person approval (4-eyes)
When it is on, kills, bans and lockdowns an operator starts need a second approver, who must give a reason; automatic responses follow your autonomy level. Nobody can approve their own request.
Enterprise LIVESSO (OIDC / Google Workspace)
Google Workspace, or any OpenID Connect provider (Okta, Entra ID, Auth0, Keycloak): code flow with PKCE, tokens checked against the provider's published keys. Domain hint and auto-provisioning supported.
Enterprise LIVESSO (SAML 2.0)
SP-initiated SAML 2.0 for Okta, Entra ID, and any SAML 2.0 IdP. Signed assertions required, clock-skew tolerance, auto-provisioning.
Enterprise LIVESCIM 2.0 user provisioning
SCIM 2.0 user create, update, deactivate, and delete, group-to-role sync, and discovery endpoints. Works with Okta and Entra ID.
EnterpriseIntegrations
4 LIVE LIVEWebhooks
Send events to any HTTPS endpoint, HMAC-signed, with rotatable secrets.
Business+ LIVEAPI keys (scoped)
Keys scoped to read or write access per area of the product. Shown once.
All plans LIVESlack, Teams, PagerDuty, Jira and more
Send alerts to Slack, Microsoft Teams, PagerDuty, Opsgenie, Discord, Jira, ServiceNow, and others. Slack messages carry Approve/Reject buttons for approval requests. Investigations can be linked to an existing Jira issue.
Business+ LIVESplunk and Datadog
Send alert events to Splunk (HTTP Event Collector) or Datadog.
Business+Automation
6 LIVE LIVEResponse pipeline
Watch, Investigate, Decide, Execute, Reflect: every automated response goes through the same stages, scored for confidence and recorded.
All plans LIVEAutopilot mode (confirmed threats auto-act)
Confirmed threats are acted on without waiting. Uncertain ones wait for a person, and isolation or lockdown from an investigation plan always does.
Business+ LIVEAction history
The rationale, confidence, signals, and outcome behind every automated decision, and any override.
All plans LIVEOverride learning
When operators override an automated decision, the correction is recorded and changes how confident Vigilhold must be before acting that way again.
All plans LIVEResponse simulator
Run an attack scenario through the response pipeline as a dry run to see what Vigilhold would do, without touching the server.
Enterprise LIVEZero-trust process network policy
Define which processes may make which network connections. Agents enforce the policy on the host and keep it across restarts.
EnterpriseFleet & Platform
3 LIVE LIVEAgent update rings
Put each server in an early, standard, cautious, or paused ring. A new agent release reaches later rings only after earlier hosts run it healthily, and a release any host failed to install is held.
All plans LIVEPush notifications
Push alerts to the Vigilhold iOS app and to your browser. Subscribe in Settings → Notifications.
All plans LIVEPublic status page
A public status page for your customers at a Vigilhold address.
Business+This page reflects the current shipping state of Vigilhold features. BETA and EXP features are functional and available to customers but may change before reaching stable. Roadmap items have no guaranteed ETA. If you are evaluating Vigilhold for a compliance-sensitive use case, please confirm the relevant features are marked LIVE before completing your procurement. Contact sales with questions.
← Back to homeTrust & security →