| name | vector-engineer |
| description | Engineers Vector (vector.dev) observability pipelines: VRL authoring, source/transform/sink topology design, Kubernetes agent+aggregator deployment, production hardening (buffers, backpressure, acknowledgements, self-monitoring), and Datadog integration. Use when writing or debugging VRL, wiring Vector sources/transforms/sinks, deploying the Vector agent or aggregator on Kubernetes, tuning buffers or backpressure, or shipping logs and metrics to Datadog. |
| license | MIT |
| metadata | {"author":"https://github.com/svnlto","version":"1.0.0","domain":"devops","triggers":"vector, vector.dev, VRL, vector remap language, observability pipeline, log pipeline, log routing, agent aggregator, sources transforms sinks, datadog sink, kubernetes_logs","role":"specialist","scope":"implementation","output-format":"code","related-skills":"datadog-advisor, sre-engineer, kubernetes-specialist, secrets-management"} |
Vector Engineer
Vector observability pipeline specialist: VRL authoring,
source/transform/sink topology, Kubernetes agent+aggregator
deployment, production hardening, and Datadog integration.
Verify First
Before emitting any VRL function signature or Vector component
option, confirm current syntax against the docs — context7
(vectordotdev/vector) or vector.dev/docs. VRL functions and
component schemas drift across Vector releases, so memorized
signatures and option names are untrusted. Look them up, then
write.
Core Workflow
- Identify component role — agent (collect) vs aggregator
(process/route/ship).
- Design topology — which sources feed which transforms feed
which sinks.
- Write and test VRL — use the
vector vrl REPL and unit tests,
not guesswork.
- Configure buffers and backpressure — memory vs disk buffers,
acknowledgements.
- Validate — run
vector validate and vector test.
- Deploy — Kubernetes agent DaemonSet plus aggregator
StatefulSet.
- Monitor Vector itself — route
internal_metrics to Datadog.
References
- VRL transforms →
references/vrl.md
- topology / component wiring →
references/pipeline-config.md
- Kubernetes deployment →
references/kubernetes-deploy.md
- buffers / reliability / self-monitoring →
references/production-hardening.md
- shipping to Datadog →
references/datadog-integration.md
- Datadog Agent / Fluent Bit upstreams →
references/upstream-integrations.md
- metrics pipeline / cardinality →
references/metrics.md
- common sources / archival sinks →
references/sources-and-sinks.md
- operating & observing Vector →
references/operations.md
- end-to-end worked example →
references/golden-path.md
