Skip to content

jevernetes

Rust CLI and TUI that streams Kubernetes logs, applies offline rules and Jev analysis to flag what matters, then hands off.

ProjectbetaMIT

Overview

Jevernetes collects and analyses Kubernetes logs with a single Rust binary. It provides bounded streaming ingestion, local offline rules, typed Jev analysis that decides which lines matter and what to investigate next, an interactive terminal UI and JSON reports, with agent handoff for deeper investigation. Collection is read-only and judgments are advisory.

It uses the Rust Kubernetes client with your kubeconfig or an in-cluster service account and needs no kubectl or Python. Offline file analysis makes no network requests; online analysis requires a TypeSafe key supplied through an environment variable or key file, never on the command line. Live mode reports queue pressure and coverage gaps, batches analysis, and saves a final report on interruption or when a duration or budget limit is reached, with exit code 2 meaning partial coverage. The TUI has Important, Routine, Needs Review, All and Search tabs, lets you ask Jev a question about the retained window, and keeps completed collections open for browsing. It is MIT licensed and verified on Rust 1.94.

What the decision model decides

TypeQuestionAnswers
ChoiceIs this log line important, routine or in need of review?(inferred)Important · Routine · Needs Review
ChoiceWhat should be investigated next?(inferred)—

How it works

  1. Stream logs from Kubernetes or files
  2. Apply offline rules
  3. Batch lines to Jev for typed analysis (online mode)
  4. Show results in TUI or JSON report; hand off to an agent

Strengths

  • Read-only and advisory
  • Offline mode with no network
  • Budget and duration limits with partial-coverage reporting

Caveats

  • Online analysis sends log text to TypeSafe
  • Judgments are advisory

Getting started

cargo build --release --locked, then `jevernetes k8s -f --tail 0 --tui` (add --offline for no network).

At a glance

Type
Project
Status
beta
Language
Rust
Runtime
CLI / TUI
Has code
yes
License
MIT

Use cases and integrations

How curated lists describe it

  • Live Kubernetes log analysis, contextual investigation, and agent handoff powered by Jev.

    logicrw/awesome-jev-projects
  • Live Kubernetes log analysis, contextual investigation, and agent handoff powered by Jev.

    hellogumbo/awesome-jev
  • Reads Kubernetes logs, decides which lines matter and what to investigate next, and hands off to an agent. Rust, with an offline mode.

    MrJev/awesome-jev

Frequently asked questions

Does it change anything in my cluster?

No, collection is read-only and judgments are advisory.

Can it run without a TypeSafe key?

Yes, offline mode uses local rules only.

What does exit code 2 mean?

Partial coverage, not a crash.

Related projects

  • Scores OpenTelemetry logs with Jev for diagnostic value, priority and routing before any expensive LLM analysis.

    ProjectbetaTypeScript
  • Splunk search command that asks Jev typed questions about events and adds probabilities, choices and scores as fields.

    AppbetaPython
  • Smart-meter fleet dashboard where code pre-filters telemetry and Jev classifies anomalies for remediation across millions of endpoints.

    ProjectprototypeTypeScript
  • OpenTelemetry Collector processor that asks Jev how useful each metric is from its metadata and applies keep/drop policy.

    ServiceprototypeGo
  • OpenTelemetry trace processor that asks Jev whether an operation is diagnostic, business-critical and worth keeping, and annotates spans.

    ProjectprototypeGo
  • Industrial demo replaying compressor telemetry where code flags anomalies and Jev picks the manual fault, severity and certainty.

    DemoprototypePython