What does the | jev search command do?

It asks TypeSafe Jev typed questions about your events and adds the answers as fields. Each question is noul (is this true), choice (which of these), or score (how much), and the resulting field arrives in the same search result stream.

How do the question types map to SPL?

The README's table: noul returns a probability that SPL can sum, making the sum an expected count; choice returns one of your options plus confidence for grouping or joining; score returns a position on levels you describe for ranking.

What are the installation requirements for Jev for Splunk?

Splunk Enterprise 9.3 or later, the KV store enabled on the search head, outbound HTTPS from the search head to api.typesafe.ai, and a TypeSafe API key. Indexers and forwarders don't need the app; it runs only on the search head.

Is my event data sent to TypeSafe?

Only when someone runs | jev: it sends just the fields a search names and the question to api.typesafe.ai, billed to your TypeSafe account. Nothing leaves Splunk before that, and the cache stores answers without the event text.

How is caching handled?

Answers are cached in the KV store, one entry per model, question and state (the named fields). Asking the same question again, or from a dashboard, only pays for events that were never judged.

How do I install and check it?

Download the release tarball from GitHub and install it on the search head via Apps > Manage Apps > Install app from file, then restart Splunk. Open Jev for Splunk > Setup, paste your key, run the self-test, or check with the jevtest live=true search.

What permissions do users need?

Give the people who run | jev the jev_user role. The README says they do not need the list_storage_passwords capability, so the API key can stay restricted to whoever runs setup.

Jev directory

CommunityJev Integrations & Adapters

Jev for Splunk

A Splunk app whose | jev search command asks TypeSafe Jev typed questions about your events and adds calibrated probabilities, choices or scores as fields, cached in the KV store so you pay only for unjudged events.

See highlights Open github.com

Stars: 1
Language: Python
Reading time: 3 min
Published by: Community
Author: sispehar
Last verified: Oct 2026

sispehar/jev-for-splunk

User repository on GitHub

View on GitHub

Splunk app: the | jev search command asks TypeSafe Jev typed questions about your events and adds calibrated probabilities, choices and scores as fields, cached in the KV store.

Forks: 0
Last pushed: Updated Sep 2026

Repo stats from the GitHub API, cached Oct 2026.

Tagscommunity catalog-import logicrw splunk spl search

Works forJev for Classification Jev for Scoring

Used inJev for Security & Privacy Jev for Customer Support

Languages: Python
Capabilities: Self-hostedRequires API key

Highlights

Three question types map to SPL: noul returns a probability you can sum, choice returns one option to group or join on, score returns a level to rank by.
A question can replace piles of regexes: the README finds Bash commands passing secrets to plink without a pattern for each tool.
Answers cache in the KV store per model, question and named fields, without the event text, so dashboards pay only for events never judged.
Runs on the search head only, sending just the named fields and the question to api.typesafe.ai, billed to your TypeSafe account.
Needs Splunk Enterprise 9.3+, the KV store, outbound HTTPS and a TypeSafe key; the jev_user role avoids list_storage_passwords.

Quickstart

bash

./scripts/build.sh --appinspect   # dist/jev_for_splunk.tar.gz

Watch out

Grounded in the truncated README and GitHub metadata reviewed on 2026-10-15; the README was cut off in the install-check table, GitHub reports the licence as NOASSERTION, the app is not on Splunkbase and shares its folder name with a different unofficial app, and results depend on your fields and thresholds.

Questions about this entry

Short answers drawn from the linked project or page.

Reactions & coverage

Posts, threads, and videos about this entry from around the web.

X: LinkedIn bookmarks search

X: 1,066 reaction videos, searchable

X: Intent-based search in a Chrome extension

X: Margin, bookmark search

From the Jev guides

Original write-ups on what this entry is for.

Decision gates in agent frameworks: what each integration ships

Oct 16, 2026

How Pydantic AI, Mastra, TanStack AI, DSPy, Spring AI, and LiteLLM expose Jev: where the typed decision sits in the loop, what each integration returns, and which big frameworks are only adjacent.

Jev Integrations & Adapters#agents#frameworks#integrations

Read article

From the Jev AI model community

Posts from builders shipping with Jev right now.

Follow @typesafeai