Why split deterministic checks from Jev judgment calls?

Exact-match rules stay in plain Python regex for free, and only semantic questions like positioning contradictions or out-of-context quotability go to Jev in one batched call per page.

How much brand setup is needed before the first scan?

You fill src/brand_rules.py with positioning, retired terms, approved numbers, and known mistakes, ideally by pasting your style guide into a coding assistant.

Do I have to approve a pilot before scanning the whole site?

Yes. You build page lists from your sitemap, run the pilot, review the dashboard and JSON, then create pilot_approved.txt because run_full.py refuses without it.

What happens if a thousand-page run dies halfway?

Every result is cached to disk keyed by URL hash with atomic writes, so a re-run only processes missing pages and never pays twice for a scored page.

What did the real-world run look like?

1,256 pages, 0 errors, about 6 minutes and $0.16 total. Each page gets one batched Jev call costing about $0.00002 to $0.00003, while the deterministic checks make no API call at all.

What does it check on every page?

Free deterministic checks cover retired terminology, unapproved numeric claims, filler and hedge words, known copy-versus-brand mismatches, plus structural and AEO checks such as schema.org presence, paragraph and bullet usage, heading hierarchy, an author byline and a visible updated or published date. Jev adds semantic judgments such as whether the positioning contradicts your brand statement.

What does the output look like?

Every run writes full structured results as JSON, a flattened CSV with one row per page ready for LLM tiering or auto-drafted fixes, and a self-contained offline Plotly dashboard showing score distribution, category failure breakdown, a staleness-versus-score scatter and a worst-offenders leaderboard.

CommunityJev Integrations & Adapters

Jev Site Auditor

Site-wide content auditor combining free deterministic regex and structural checks with one batched Jev judgment call per page.

See highlights

Open github.com

Stars: 0
Language: Python
License: MIT
Published by: Community
Author: tjkimcloud
Last verified: Oct 2026

tjkimcloud/jev-site-auditor

User repository on GitHub

View on GitHub

Score every page on your site against your own brand/AEO rules, combining free regex checks with Jev (TypeSafe AI) semantic judgment calls. Point it at your own site with your own brand rules and OpenRouter key.

Forks: 0
Last pushed: Updated Sep 2026

Repo stats from the GitHub API, cached Oct 2026.

About this project

Score every page on your website against your own brand/AEO (Answer Engine Optimization) rules, combining free deterministic checks with semantic judgment calls from Jev (TypeSafe AI's "System One" decision model) - fast and cheap enough to run against an entire site. Real-world run: 1,256 pages, 0 errors, 6 minutes, $0.16 total. For every page on your site, it runs:

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

Tagscommunity github

Works forJev for Routing Jev for Scoring

Used inJev for Software & DevTools

Languages: Python

Highlights

Runs deterministic brand, numeric, filler, schema.org, heading, byline, and date checks with no API call.
Batches semantic brand, claim, pricing, and answer-engine judgments into one Jev call per page.
Produces a composite 0-100 score across 8 weighted categories with per-category reasons.
Writes JSON plus flattened CSV plus offline Plotly dashboard with resumable disk cache.

Watch out

Brand rules need manual calibration; the full run is gated on approving a pilot so a bad rule is caught on a small sample. Copy reviewed against the repository README on 2026-10-04, not from running it.

Questions about this entry

Short answers drawn from the linked project or page.

More Jev resources like this

AutoGPT

187.5k

Autonomous agent platform with 187,504 stars for building and running agents from plain English or a visual builder. Its platform features first-class TypeSafe Jev decision blocks for typed routing, filtering, scoring and confidence-gated next-action dispatching.

Jev Integrations & Adapters#community#autonomous-agents#platform

Community

Significant-Gravitas/AutoGPTPython

AgentScope

33k

Alibaba's multi-agent platform, which implements native TypeSafe Jev classification through JevClassifierModel for binary, choice and score routing across agent pipelines. Python 3.11+, Apache-2.0, with a FastAPI agent service and web UI.

Jev Integrations & Adapters#community#multi-agent#framework

Community

agentscope-ai/agentscopePython

JIT-JEV Context OS

Epistemic runtime for autonomous coding agents that uses Jev as a sub-100ms System 1 gate to shadow-evaluate capsule facts, propose evictions before bloat and prefetch memory between turns. Its fixture run reports 31.3% fewer turns and 52.6% fewer blind discovery operations, with a fail-open circuit breaker.

Jev Integrations & Adapters#community#agents#context-engineering

Community

wojciechwiesner/jit-contextPython

ai-chess-lab

Browser chess lab where LLMs, System One models, and humans play single games, Swiss tournaments, and leagues with Stockfish analysis and cost tracking.

Jev Integrations & Adapters#community#github

Community

dperezcabrera/ai-chess-labPython

typesafe-jev-plugin

Claude Code plugin that turns classification problems into reusable .sig.json signatures, then scores CSV, Excel, or text datasets row by row through the Jev API and writes results back to Excel.

Jev Integrations & Adapters#community#github

Community

arnab621/typesafe-jev-pluginPython

systemone

Compatibility server exposing the /v1/systemone typed-decision API on top of vLLM, SGLang, or OpenRouter models, with vision models enabling image inputs that Jev and Laya lack.

Jev Integrations & Adapters#community#github

Community

infinitylogesh/systemonePython

Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai