Run open-jev- and Laya-shaped typed-decision models in the browser: one state plus typed questions in, calibrated probabilities out.

JSON becomes real things. Define your catalog, register your components, let AI generate.

Connect Claude Code, Codex, or Hermes on your computer to a thinkpool Code room.

Local-first Nitro server for the Tessera database analysis workspace on Node.js and Bun.

Jev validation guard for pi: on-demand tool + automatic provider gate. Dual backend: TypeSafe direct + OpenRouter.

A pi extension that uses TypeSafe AI's Jev model to curate context before each LLM call, keeping only the context the next call is likely to need.

MCP server exposing TypeSafe Jev as purpose-built judgment tools: verify claims, screen for injection, find, rerank, classify, decide, compare passages, extract fields, review patches, and gate completion claims, each with typed probabilities.

Local-first observation and heuristic risk scanning for AI coding agents (Claude Code, Codex, Cursor, Copilot CLI, Gemini). Redacts secrets, flags risky tool calls, scans skills and MCP configs offline.

Claude Code / Codex / pi plugin that hands the agent steps needing no text output to Jev (TypeSafe's judgment model) — measured p50 ~230 ms and ~$0.02 per 1,000 judgments, with typed escalation back to the LLM.

Configurable Jev browser research and classification CLI

Code quality and coverage for coding agents

Community fork of Luke Parke's pi-subagent with Jev model/tool routing and verified Pi child capabilities

Clearer Pi replies with Jev review, optional rewrite/visuals, and on-demand bilingual settings.

TypeSafe Jev fast-path decision provider, dynamic action space and decision validation.

A Jev-driven browser agent as an MCP tool: give it a task and a start URL; it navigates with typed Choice actions, a typed-input cascade, and stop/stuck safety gates.

Gives DeepSeek Harness three automatic Jev judgments — prune oversized tool output, screen fetched pages for injected instructions, suggest which skill to use — plus the jev_ask and jev_gate tools, a persistent judgment ledger, and a /jev-status diagnosis

A natural-language linter: ast-grep picks the code, a sentence you write is the rule, and Jev returns the verdict

Interactive browser terminal and automation CLI powered by Jev

Local LLM gateway that lets TypeSafe's Jev model pick the tool for coding agents. Launchers for Codex, Claude Code, OpenCode and Gemini, with a monitoring dashboard.

Developer Harness with TypeSafe AI / Jev for AI Coding Agents