Watch TypeSafe Jev play Slay the Spire 2, with a local decision dashboard and offline evaluations.
Local & offline models — page 8
Local alternatives and on-device decision models that follow the System One interface, plus shims for existing Jev clients.
368 active projects match this task, ordered by traction.
Skill Dash uses Jev to judge Claude Code and Codex skills and plugins: usefulness, redundancy, clarity, duplicates, safety. Local dashboard plus the corpus pipeline behind whichskills.dev. MIT.
Windows port of typesafe-computer-use with Jev decisions, local OCR/UIA, and MCP host integration
A local proxy that sits between Claude Code and the Anthropic API and uses TypeSafe's Jev to route each subagent to the model the task actually needs. Your main conversation keeps whatever model you picked
1
Pre-registered benchmark: can a 2B local model (Gemma 4 E2B) answer web questions without making things up when a decision model (TypeSafe Jev) makes every call? SearXNG for search, MemPalace for verbatim memory, seven arms including open local judges. Spec and thresholds fixed before any run.
1
Instant semantic code search: local ripgrep shortlist, meaning-ranked by TypeSafe Jev via OpenRouter.
1
Jev-powered Judgment layer for Claude Code. Stops paying reasoning prices for if-statements: a PreToolUse hook scores every tool call against a YAML policy you own — allow / deny in ~100 ms, no LLM in the loop. TypeSafe Jev now, local models next. Dry-run by default, calibration table published.
Standalone TypeSafe Jev code-review CLI: typed decisions over a local git diff.
1
Local evaluation workbench for TypeSafe Jev
Local stand-ins for TypeSafe Jev's System One interface: typed calibrated decisions from small models, with a routing benchmark and a distilled 19 ms encoder
1
A gate for your agent's expensive steps, powered by TypeSafe Jev (System One). Offline-first, OpenRouter or direct, MIT.
Rust CLI powered by Jev from TypeSafe.ai that ranks agent skills for the next step using live session context. Includes Claude Code hooks, structured JSON, abstention, and local feedback. Requires a TypeSafe API key.
Local browser automation UI that uses TypeSafe Jev to choose bounded page actions and a text model only for field values.
Typed, auditable Jev decisions for Python agents, with an optional shared runtime, offline previews, and calibration.
Application of TypeSafe Jev (noul judgment primitive) on the collusion.wiki corpus: agent vs human page authorship, head-to-head vs local Qwen3.8-Flash-Next
Typed Elixir client for TypeSafe AI and its Jev System One model, with offline test stubs, concurrent fan-out, and atom-keyed answers.
ForgeOS Browser - a private, local-first browser built for humans and prepared for trusted AI agents. Educational prototype. Zero telemetry.
AgentPulse is a dynamic API reference for local agents.
1
Tiny Rust LLM proxy: one local endpoint over many providers, auto-routing with quota-aware fallback, launches claude/opencode/pi
NEXUS is a local command deck that puts any mix of CLI agents — Claude Code, Codex, DeepSeek Harness, OpenClaw, or your own — behind a single cyberpunk WebUI: broadcast or @-target instructions, watch every agent in its own live window, let agents dispatch each other, and share a collective memory across the team.
Open-source macOS activity history: local-first Computer History, Screen Time and AI conversation indexing with verifiable selective sharing.
1
Local-first stateful LLM co-writer harness
The missing GUI for Claude Code - local-first dashboard for sessions, tokens, context windows, memory, skills, hooks, and MCP servers. Zero config, reads ~/.claude directly.
1
InferKit - An easy and advanced Apple inference toolkit - swappable backends (CoreML, MLX, Remote, Foundation Models), 119+ common on-device models. ObjC & Swift