Platform Engineer: AI Health Checks for Wazuh
On Sale
$67.99
$67.99
Quiet does not mean safe. Make sure Wazuh is working
The worst Wazuh outages are silent: a full disk, the shard limit, a stalled pipeline, a failing integration. It checks all of it and tells you what is wrong and how to fix it.
When Wazuh breaks, the dashboards just go quiet
Today: you find out weeks later
- A full disk or the shard limit stops new alerts
- A failing integration drops alerts without a sound
- Nobody reads the manager log until something burns
With the Platform Engineer: a daily check
- Indexer, pipeline and manager checked, worst first
- Days of disk runway and shard headroom, before they run out
- Errors grouped, with a likely cause and a fix
A Wazuh admin on call every day
- Health check: Cluster colour, disk vs the watermarks, JVM heap, shards vs the limit, unassigned shards and retention, worst first
- Pipeline: How old the newest alert is, so a stalled pipeline never looks like a calm day
- Manager: Core daemons, analysisd drops and queue usage, remoted and config validation
- Failing integrations: Alerts that are not being forwarded, and how often
- Storage and capacity: Size per index family, daily growth, days of runway and the retention that would fit
- Ingestion: Alerts per hour with gaps, noisy agents, top rules, and events received, processed and dropped
- Log triage: Errors and warnings grouped, with a likely cause for known patterns
- Upgrade check: The manager vs the latest Wazuh release, agent versions and unsupported combinations
- Fixes with approval: Replicas to 0 on a single node, a retention policy, or old alert indices deleted, all in dry run first
One command, about ten minutes
- Install: Run one command on Linux, macOS or Windows. It deploys n8n too if you need it
- Connect: Add a Wazuh indexer user with cluster monitoring, a read-only Manager API user, your chat app and an AI model key
- Ask: Ask if Wazuh is healthy. The daily platform check starts on its own
Careful with the platform your SOC depends on
- Approval every time: Cleanup needs an Approve click and starts in dry run
- Says what is reversible: Each change explains whether it can be undone before you approve it
- Recent alerts protected: It never deletes alerts younger than 30 days by default
- Honest diagnosis: Statuses come from the indexer and the Manager API. Likely causes are labelled as hints
- Data, not instructions: Log lines and index names are never treated as commands
- Short when healthy: Three lines on a good day, the full story when something is wrong
What you need
- n8n: n8n 2.34 or newer with Agents, or let the installer deploy it
- Indexer: A user with cluster monitoring and read access. Index management only if you allow fixes
- Manager API: Read access is enough
- Chat app: Telegram, Slack, Discord or Linear. A Telegram bot takes a minute with @BotFather
- AI model: An API key from any provider in the n8n Agent model picker
What you get
- 7 tool workflows and the daily platform check
- The agent with 5 skills: daily check, storage and retention, pipeline troubleshooting, log triage and fixes with approval
- One-command installer for Linux, macOS and Windows
- Setup guide and example conversations
- Single-organization commercial license
- 7-day money-back guarantee
Full details and a demo: neetrox.com/ai-agents/platform-engineer