aaddrick / building-with-typesafe-jev
git clone https://github.com/aaddrick/building-with-typesafe-jev.git
Unofficial skill that teaches coding agents to build with TypeSafe AI's Jev: typed decisions, calibrated confidence, and prior art from 150+ community projects.
+1
stars, 24h · #16,300
+~2
stars, 7 days · #25,408
+145
stars, 30 days · #5,184
145
★ total · +1% this week
11
forks · 1 open issues
Star history (as recorded here)
LanguagePython
LicenseMIT
Created2026-09-26 (27d ago)
Last push22d ago
Watchers0
OwnerUser
GitHub topics
agent-skillagent-skillsai-agentscalibrationclassificationclaude-codeclaude-code-pluginclaude-code-skillclaude-code-skillscodex-cligemini-cliguardrailsjevllmllm-toolsprompt-engineeringstructured-outputsystem-onetypesafetypesafe-ai
Links
- GitHub: github.com/aaddrick/building-with-typesafe-jev
- Homepage: https://github.com/aaddrick/building-with-typesafe-jev#readme
- Releases · Issues · Hacker News mentions
README excerpt
Building with TypeSafe Jev Typed decisions with calibrated confidence, for your coding agent. Links to 150+ community projects, sorted by shape, with a code sketch for each. Connect on LinkedIn! English · 简体中文 · 日本語 · 한국어 · Tiếng Việt · Português (BR) · Italiano > [!NOTE] > This is an unofficial, community skill. It is not made, reviewed, or endorsed by TypeSafe AI. TypeSafe publishes its own skill at typesafe-ai/skills. See How this differs from the official skill. Coding agents treat Jev like one more chat model. This skill teaches them to design for it: typed questions, calibrated confidence, and links to 150+ community projects, sorted by how they work, with a code sketch for each pattern. It installs in Claude Code, Codex, Antigravity CLI, Muse, and Muse Code. Jev is a System One model. It does not write text. You send it content and a set of typed questions, and it answers each one with a value and a calibrated probability, usually in 100 to 200 ms: - **Choice** picks one option from a list. Example: route a ticket to billing, shipping, or support. - **Score** places the content on a scale you describe. Example: grade a pull request from "ignores the spec" to "meets the spec". - **Noul** gives the probability that a yes/no statement is true. Example: "this shell command deletes files outside the project." ## Install Claude Code [code] [code] The skill loads on its own when you work on Jev code. To load it by hand, type: [code] Claude Desktop, Cowork, and claude.ai **Step 1.** Open **Customize > Plugins**, select **Add**, then **Add marketplace**. **Step 2.** Choose **Add from a repository**. **Step 3.** Enter `aaddrick/building-with-typesafe-jev`. Leave **Sync automatically** on so the plugin updates when this repository does. Then select **Sync**. **Step 4.** Select **Add** next to **Building with typesafe jev**. **Step 5.** Claude confirms the plugin is installed. It also appears in the desktop app and Cowork on the same account. Codex [code] [code] Start a new thread. Codex loads the skill when the task matches. To load it by hand, type: [code] Antigravity CLI [code] Check that it installed: [code] Start a new session. Antigravity CLI loads the skill when the task matches. To load it by hand, type: [code] Coming from Gemini CLI? If `agy plugin import gemini` brought this extension over, run the install command above anyway so the current copy replaces the imported one. Muse (muse.ai) Muse loads skills from `~/workspace/skills/` on its own computer. Paste this command into a Muse chat and ask Muse to run it: [code] The script copies the skill folder there and rewrites the `SKILL.md` header into the shape Muse reads. Start a new chat. Muse loads the skill when the task matches. To update, run the command again. Muse Code Clone the repository: [code] Install the skill for every project: [code] Check that it installed: [code] Start a new session. Muse Code loads the skill when the task mat