Browser automation with decision models

Reproducible evaluation harness and field note for the Jev Ultrafast research browser, with QC-graded baseline and stress suites.

Open 2B browser-agent decision model fine-tuned from MiniCPM5 that serves the System One protocol for jev-ultrafast.

Pre-action gate and browser agent where rules decide first, then Jev or a local Jev-compatible model, emitting allow, confirm or block.

Local social research agent: Jev picks each bounded Instagram, TikTok or LinkedIn step and socai CLI runs it in your Chrome.

Browser extension and MCP server where Aside observes the page and Jev picks the next action from defined candidates.

Python library for typed decisions with Jev or your own provider, adding validated distributions, batching, budgets, ranking and records.

Playwright plus Jev core behind a typed SDK, persistent CLI and MCP server, with native operations needing no key and verifiable assertions.

Library, CLI and MCP server where an LLM states each step's outcome and Jev decides element, action, value and done/blocked/irreversible.

Reproducible bake-off of Playwright CLI alone, CLI plus Jev, and Playwright MCP on fuzzy browser goals, with honest results.

Zero-build local playground for Jev with browser-agent, intent routing and context-pruning presets, including a Playwright agent.

Evidence-driven frontend QA where Jev picks browser actions while code checks DOM, HTTP and database contracts.

Browser agent that acts on one look at the page with one open-model call per action; Jev is an optional chooser backend.

Coding-agent skill that drives Chrome over CDP to hunt for jobs on company sites, with Jev scoring every link and job posting.

OpenJev 1.7B-8B: open decision models distilled for general decisions and browser-agent steps, served behind /v1/systemone.

OpenJev on Hugging Face: open-weights decision model for text, web pages and screenshots with up to 52 options per forward pass.

Open 2B non-autoregressive decision model on Qwen3.5-2B for agent tool routing and DOM actions, with BF16, FP8 and INT8 weights.

706K-parameter Jev-style option scorer that rates fill, check, click or skip for each form element in one parallel pass.

Open stack to synthesise decision data, train and calibrate a Qwen3.5-0.8B scorer, evaluate on Mind2Web and serve Jev-style.

Open Apache-2.0 browser-agent decision model (Qwen3.5-35B-A3B fine-tune) that doubled Jev 1.13's task success in the same agent.

Browser driver where Jev makes every decision and agent-browser performs every action, benchmarked against Claude Code on 10 tasks.

Jev-powered browser MCP server that makes each step in one parallel call and escalates to the driving LLM only when unsure.

Browser and macOS desktop agent for Pi where Jev picks each action from a structured text observation in a bounded loop.

Single-file, zero-dependency port of jev-ultrafast to Ego Lite: one typed-choice Jev request per browser step, with extra guardrails.

Runs jev-browser entirely on a local Jev-style engine (Qwen2.5 1.5B decisions, 0.5B typing) on an RTX 2060 SUPER, with no cloud API.

Browser-agent benchmark comparing WebMCP with screenshots, DOM and code execution, including a Jev + Mercury 2.5 configuration.

Drives ego-browser pages with Jev: code builds the allowed actions, Jev picks one, and code acts and re-checks.

X preview of fast browser automation using Jev with OpenCode's browser CLI.

X post claiming Playwright CLI with Jev choosing each step was 98% cheaper and twice as fast as the Playwright MCP.

X post where the accessibility tree is the state, candidate actions are the options, Jev picks each step and Stagehand executes.

X demo where Jev picks which candidate link to click on a page to move toward a goal.