A curated list of resources, projects, tools, examples, research, and things people are building with Jev by TypeSafe.
Jev is a decision model from TypeSafe designed for making bounded, structured decisions rather than generating open-ended text.
This repository collects interesting things built with Jev — from GitHub projects and demos to integrations, experiments, articles, research, and community resources.
- Official Resources
- GitHub Projects
- Applications & Use Cases
- Integrations
- Demos & Experiments
- Articles & Blog Posts
- Research & Papers
- Videos & Talks
- Tutorials & Examples
- Community
- Other Resources
- Contributing
Resources maintained by or directly related to TypeSafe and Jev.
- TypeSafe — Official TypeSafe website.
- Jev — Learn more about Jev.
- Documentation — Official Jev documentation.
- GitHub — Official Jev repositories.
Add official documentation and repositories here as they become available.
Open-source projects and repositories that use, integrate, experiment with, or demonstrate Jev.
- jev-ultrafast — Browser agent that uses Jev to select browser actions and only calls a text model when text generation is needed.
- fast-jev-compaction — Claude Code plugin and npm library using Jev to make context-compaction decisions while preserving important tool-call context.
- jev-trader — Trading bot experiment using Jev to make trading decisions on Monad.
- tax-doc-classifier — Tax document classifier using Jev to classify IRS tax forms.
- jev-drone — Simulated drone using camera input and Jev for real-time decision making.
- killmyidea — Startup idea evaluation tool that uses Jev to assess ideas and return structured decisions.
- jev-curate — High-throughput data curation tool that uses typed Jev judgments to filter Parquet and JSONL data.
- pg-jev — PostgreSQL extension that uses Jev to answer natural-language questions about database tables.
Examples of real-world problems where people are using or experimenting with Jev.
- 🤖 AI Agents — Agent routing, tool selection, escalation, and decision making.
- 🔀 Model Routing — Selecting models based on the task or context.
- 🛡️ Guardrails — Deciding whether an AI-generated action should be allowed.
- 📚 RAG — Retrieval relevance, answer verification, and routing.
- 🔐 Security — Risk classification and security decisions.
- 💳 Finance — Risk assessment, transaction classification, and fraud-related decisions.
- 📄 Document Processing — Document classification and workflow routing.
- 👨💻 Developer Tools — Coding agents, code review, issue classification, and automation.
- 🎮 Games & Simulations — Using Jev to make bounded decisions inside games and simulations.
Projects that connect Jev with other technologies and frameworks.
- LangChain
- LangGraph
- LlamaIndex
- PydanticAI
- CrewAI
- OpenAI Agents SDK
- Google ADK
- FastAPI
- Django
- Flask
- Pydantic
- MCP servers
- MCP tools
- Agent + MCP integrations
- Redis
- Kafka
- PostgreSQL
- Docker
- Kubernetes
Interesting experiments, prototypes, and proof-of-concepts built with Jev.
- Project Name — Short description.
Have you built something unusual with Jev? Contribute it!
Contributions are welcome! 🎉
If you have built something with Jev or found an interesting Jev resource, please consider adding it.
You can contribute:
- 🐙 GitHub repositories
- 🧩 Libraries and tools
- 🤖 AI/agent projects
- 🔌 Integrations
- 🎮 Demos
- 📝 Articles
- 📚 Tutorials
- 🔬 Research
- 🎥 Videos and talks
- 📊 Benchmarks
- 💡 Interesting experiments
- Fork this repository.
- Add your resource to the appropriate section.
- Keep the description short and factual.
- Use the existing formatting.
- Submit a pull request.
- **[Project / Resource Name](URL)** — Short description of what it does and how it relates to Jev.Please avoid promotional descriptions, duplicate entries, and projects that have no meaningful connection to Jev.
If you find this collection useful, consider giving the repository a ⭐ and sharing it with other developers interested in Jev.
Built and curated by the python-world team 🐍
This list is available under the MIT License.