タグ：ai-safety

ai-safety タグの付いた Jev GitHub プロジェクトを検索・並べ替えし、元のリポジトリを確認できます。

194jev-harness
A custom coding harness for TypeSafe AI's Jev: an LLM proposes, Jev answers narrow questions, code decides, every step leaves a receipt.
TypeSafeAI/jev-harness · TypeScript
3010
agent-harnessagentic-codingai-agentsai-governanceai-safetycode-reviewcoding-agents
214dsh-jev-interceptor
⚡ Millisecond System-1 judgement for every tool call in DeepSeek Harness — Jev-powered risk classification & evidence-gated auto-approval. Fail-closed by construction. dsh 生态第一个 System-1 决策插件
AskTheWay/dsh-jev-interceptor · TypeScript
243
agent-harnessagentic-aiai-agentsai-safetydecision-modeldeepseekdeepseek-harness
380jevc
Compile agent policy prose into deterministic verdict programs: narrow evidence questions for the model, the verdict computed in code. Install: npm i -g jev-compiler
doronp/jevc · TypeScript
80
agent-securityai-agentsai-safetyai-sdkclaude-codecli
398doom-or-bloom
Doom or Bloom: map your AI worldview and compare it to others.
transitive-bullshit/doom-or-bloom · TypeScript
70
ai-futuresai-risksai-safetyworldviewtypescript
550jevshield
Sub-100ms security gate for AI agent tool calls, powered by TypeSafe's Jev (System-1) decision model. Single-request Choice/Noul/Score evaluation, dual-factor blocking matrix, calibrated-confidence routing, fail-closed parsing, zero-config local fallback. LangChain-ready.
lgy1027/jevshield · Python
40
agent-securityai-agentsai-safetyguardrailslangchainllm-securitymiddleware
628jev-shield
Semantic MCP firewall powered by Jev — screens every tool call, tool result, and tool description with calibrated System One verification. 94% block recall, 0 false positives, ~$0.00002/check.
caiovicentino/jev-shield · JavaScript
30
javascriptai-safetymcpprompt-injection
689actiongate-jev
Open-source Jev tool-calling authorization gateway for AI agents: deterministic policy, exact-action single-use permits, MCP and HTTP enforcement.
omkarghugarkar007/actiongate-jev · TypeScript
20
agent-securityagentic-aiai-agentsai-guardrailsai-safetyauthorizationfastify
834tinystruct-typesafe-sdk
A tinystruct-based TypeSafe SDK with JEV model.
tinystruct/tinystruct-typesafe-sdk · Java
20
ai-safetyai-toolstinystructtypesafe-actions
991jev-guard
Vets an LLM agent's tool calls through TypeSafe AI's Jev before they run — allow, block, or hold, failing safe on uncertainty.
CMaintz/jev-guard · TypeScript
11
ai-agentsai-safetyguardrailsllmtypescript
1093Jevs-Garage
A garage full of tiny experiments for building critical systems with System One & Jev 🔧🧠⚡
JGalego/Jevs-Garage · Python
11
ai-demosai-safetydecision-intelligencedeveloper-toolsexplainable-aihuman-in-the-loopincident-response
1157skill-dash
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.
48Nauts-Operator/skill-dash · Python
10
agent-skillsai-safetyclaude-codeclaude-skillscodexdashboardprovenance
1232approval-judge-bridge
OpenAI-compatible judge endpoint for agent approval gates: typed judgements (Jev), any OpenAI-compatible model, or a rule file — fail-closed, calibrated, with a replay battery
oppih/approval-judge-bridge · Python
00
ai-agentsai-safetyapprovalsguardrailshermespython
1373foreman-jev-evaluation
JEV evaluation (FM-JEV-01): replay-only, advisory-only evaluation of Foreman-style supervision with TypeSafe Jev. Measures how much safety comes from the model versus a deterministic evidence gate. No worker authority.
MahdiHedhli/foreman-jev-evaluation · Python
00
agent-supervisionai-safetyevaluation-harnessllm-evaluationpython
1530Jev-Defense
A Jev-powered security guard for AI agents: blocks dangerous tool calls, strips prompt injection, scans skills. Works with Claude Code, Codex, Copilot CLI, Gemini CLI, Cursor, and OpenCode.
prestonkakukdev/Jev-Defense · Python
00
ai-agentsai-safetyclaude-codeguardrailsllm-securityprompt-injectionpython
1557jev-enterprise-decision-fabric
Architecture for running many semantic decisions through one validated path, with a labelled 111-case benchmark comparing TypeSafe Jev against a Claude baseline, and a dashboard for inspecting any single decision. Experimental, not production.
ghubnab99/jev-enterprise-decision-fabric · C#
00
agent-toolsai-safetyaspnetcorebenchmarkc#decision-makingdotnet
1813jev-watchdog
Event-driven trajectory watchdog for coding agents: per-event LLM judgments (Jev, Claude, GPT) accumulated into quarantine decisions
glebmish/jev-watchdog · Python
00
ai-agentsai-safetyclaude-codeclaude-code-hooksguardrailsllm-as-a-judgepython
1819jev-workflow
Decision contracts, adversarial testing, tracing, stability, and privacy controls for TypeSafe Jev
gbesse/jev-workflow · TypeScript
00
ai-safetydecision-engineobservabilityprivacytypescript
1904laya-guardrails
Guardrails de input, tool call y output para agentes de IA con un modelo System One tipo Jev (laya-pt-es-typed) autoalojado en una NVIDIA L4: 5 ms por check frente a 140 ms de un LLM-as-a-judge.
javimp2003/laya-guardrails · Python
00
ai-safetyfastapiguardrailslayallm-agentsllm-as-a-judgenvidia-l4

18 件中 1–18 件