Is the Jev rerank on by default, and what leaves my machine?

It is off until you configure it, and enabling it sends the result text to TypeSafe's servers. The default embedder is lexical and fully offline; the optional neural embedder downloads about 90 MB.

Are those token savings real or just marketing?

They come with a stated method: 8 tasks, 4 languages, 16 runs per arm, with real token counts from a headless agent. XERJ used 9,982 output tokens against 26,477 for a grep-driven agent at the same 16/16 solve rate. The README also concedes the win is gated by memorisation and can be neutral to harmful on popular public libraries.

What does --insecure actually expose?

It turns off TLS and API-key auth, so anything that can reach the port is a superuser who can read and delete every index, including the /_memory namespaces. A fresh node binds 127.0.0.1 only, and a cleartext bind on any other interface refuses to start unless you allow it.

Where is the Jev rerank actually documented?

Not in the part of the README I could read: that section covers indexing, search, the security model, and reference coding, with no rerank parameters, thresholds, or failure behaviour. The listing records that rerank stays off until configured and that the hev Python client works against a node unchanged.

What is XERJ written in, and how do I install it?

It is a Rust project under Apache-2.0 distributed as a single static binary with no JVM. You can install it with a curl script or a one-prompt agent install that points at xerj.org/llms.txt, and there is a hand-verification path for Windows and for anyone who wants to read the install steps.

What is reference coding, and what does it measure?

XERJ's headline use case: point an agent at a task and it clones and indexes the open-source repositories closest to that problem, then the agent reuses how they solved it before writing its own code. The case study measures 2.7 times fewer output tokens than a grep-driven agent at the same 16 of 16 solve rate.

What does the project ask for in return?

One field report. Running it and sending a short report of what you pointed it at, what you used it for, and your honest verdict is described as the membership card for a free Apache-2.0 binary. Reports are CLA-exempt, live under user-feedback, a negative verdict counts, and a feedback command can open the pull request for you.

CommunityJev Tools & CLIs

XERJ

An Elasticsearch-compatible autoindex search layer that optionally sends top hits to TypeSafe Jev and reorders results by relevance scores.

See highlights

Open github.com

Stars: 3.5k
Language: Rust
License: APACHE-2.0
Published by: Community
Author: xerj-org
Last verified: Oct 2026

xerj-org/xerj

User repository on GitHub

View on GitHub

XERJ is the new way for AI to search data. Its autoindex capability activates agents to know your data without the token waste of grep and sed. One command indexes code, docs, logs and PDFs for search, RAG, security audits and agent memory, using 40x fewer tokens than grep. Elasticsearch compatible, so existing clients just work.

Forks: 242
Last pushed: Updated Oct 2026

Repo stats from the GitHub API, cached Oct 2026.Homepage

About this project

XERJ is a community-trusted local AI search that indexes any folder automatically, so your coding agent stops burning tokens reading files one by one and pulls the exact code it needs instead. Reference coding is its main use case and the clearest win: point an agent at a task and it downloads the open-source repos closest to it, indexes them, and reuses how they solved the problem before writing its own code. In a controlled study that cut a coding agent's output tokens by 2.7x at the same 16/16 solve rate (case…

Quoted from the project's README. Read the full README.

Tagscommunity rust search rag

Works forJev for Ranking Jev for Classification

Used inJev for Software & DevTools Jev for Data & Analytics Jev for Security & Privacy

Languages: Rust
Platforms: API
Capabilities: Open sourceRequires API key

Highlights

README optional rerank sends result text to Jev and sorts by returned scores; stays off until configured.
Documents compatibility with the hev reranker Python client against a XERJ node without changes.
Claims 40x fewer tokens than grep for indexing code, docs, logs, and PDFs for agent memory and audits.

Watch out

Apache-2.0. Enabling rerank sends hit text to TypeSafe servers; FiQA numbers in README are author-measured.

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: Natural-language search over Zillow

X: Filtering WIP todos by meaning

X: Seven more SEO and GEO workflows

X: A keystroke oracle that reads intent in ~100 ms

Related terms

Glossary definitions related to this entry.

Line-by-line search

More Jev resources like this

GenOffice

9.2k

An open-source AI office suite whose local file search reranks the top 20 FTS hits with TypeSafe Jev relevance scores before showing results.

Jev Tools & CLIs#community#typescript#desktop

Community

genspark-ai/genofficeTypeScript

typesafe-jev crate

Rust docs for typesafe-jev 0.2.0, a blocking client that sends one state and a set of Choice, Score, and Noul questions and prices input tokens at list rate.

Jev Tools & CLIsDocs#community#docs#rust

Community

jeff

Go CLI for Jev with noul, choice, score, eval, rank, and models commands: rank sends one Score question per item per weighted YAML dimension in a single request and sums weight times score in code, while threshold flags turn a failed gate into exit code 10 for scripts and CI.

Jev Tools & CLIs#community#cli#go

Community

saembit/jeff-cliGo

jcode

20.4k

Rust coding-agent harness (MIT, ~20k stars) whose crates/jcode-base/src/jev.rs is a typed TypeSafe Jev decision transport, used for memory pruning (memory_jev.rs), browser navigation (browser_jev.rs), and voice routing (voice_intent.rs) with separate provider selectors per purpose.

Jev Tools & CLIs#community#rust#coding-agent

Community

1jehuang/jcodeRust

Skill Scanner

2.6k

Cisco's security scanner for agent skills added a System One analyzer: it sends the full skill text to POST /v1/systemone as one advisory Noul question and records the probability, while being deliberately unable to emit a finding or change a severity.

Jev Tools & CLIs#community#security#agent-skills

Community

cisco-ai-defense/skill-scannerPython

Paca

1.9k

A Go-based team workspace with optional Jev AI for agent auto-routing, task auto-fill, auto-assign, and automation condition nodes.

Jev Tools & CLIs#community#go#agents

Community

Paca-AI/pacaGo

From the Jev guides

Original write-ups on what this entry is for.

Retrieval with Decisions: Filter, Rerank, and Verify

Oct 12, 2026

Retrieval is a sequence of bounded judgments. How a decision model slots into passage filtering, semantic find, reranking, and citation checks, and what stays in deterministic code.

Jev Cookbooks#retrieval#reranking#rag

Read article

Jev on the Gateways: Vercel, Workers AI, and OpenRouter

Oct 12, 2026

The request shape barely changes across surfaces; the surrounding contract does. When to call the API directly and when a gateway, a Worker binding, or a router entry earns its extra hop.

Jev Tools & CLIs#gateway#vercel#cloudflare

Read article

Exporting Decision Models: GGUF, ONNX, and Device Runtimes

Oct 10, 2026

Taking a decision checkpoint out of Python: what the GGUF and browser WebGPU exports preserve, what they change, and the parity evidence the model cards publish.

Jev Tools & CLIs#open-models#local#onnx

Read article

Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai