When should I reach for this instead of a chat model?

Use it when code needs a snap judgment with a known answer shape such as routing, classification, scoring or guardrails. Do not use it to generate text or code or to replace a deterministic rule.

How should I phrase questions for reliable answers?

Ask atomic questions and send every independent question about the same state in one ask call, then combine answers in code. Noul 0.5 means unsure rather than medium.

How do install and credentials work across machines?

Copy the binary to a PATH directory and the SKILL.md to an agent skills host, set the key and optional base URL or model, or pass a JSON creds file instead of the env var.

Which commands cover daily checks and batch questions?

Info, models, smoke and ask cover endpoints and quick checks; ask accepts inline state or state and question files with JSON output, using Bearer auth against the evaluate and models endpoints.

What does the TypeSafe CLI depend on?

Python 3.9 or higher and only the standard library, using urllib, json and argparse. It runs on Linux, macOS or any python3, and there is no package install beyond cloning the repository and making the typesafe script executable.

Which commands does the CLI expose?

Four: info prints endpoints without the key, models lists available models, smoke runs the docs quickstart example, and ask evaluates a state against questions. ask can also read --state-file and --questions-file and emit --json.

Can a coding agent use the TypeSafe CLI?

Yes. It ships an agent skill you copy to ~/.agents/skills/typesafe/SKILL.md for Claude, Codex, Grok and other Agent Skills hosts. The full flag set and how to read answers live in skills/typesafe/SKILL.md.

CommunityJev Agents & MCP

TypeSafe CLI — geilt

Stdlib-only Python CLI and agent skill that evaluates a state against typed Choice, Score and Noul questions over the System One API without generating prose.

See highlights

Open github.com

Stars: 3
Language: Python
Published by: Community
Author: geilt
Last verified: Oct 2026

geilt/typesafe-cli

User repository on GitHub

View on GitHub

CLI and agent skill for TypeSafe System One (Jev): typed Choice, Score, and Noul judgments.

Forks: 0
Last pushed: Updated Sep 2026

Repo stats from the GitHub API, cached Oct 2026.

This is one of several Jev listings for the same project. The canonical entry is TypeSafe CLI — y0usaf.

About this project

Small Python CLI + agent skill for TypeSafe System One (Jev). It evaluates a state against typed Choice / Score / Noul questions and prints structured answers. It is not a chat model: no generated prose, no JSON-from-prose parsing. Requires Python 3.9+ (stdlib only: urllib, json, argparse). Linux, macOS, or any python3. When to use it (full writeup): USE-CASES.md. Install the agent skill (Claude / Codex / Grok / other Agent Skills hosts):

Quoted from the project's README. Read the full README.

Tagscommunity jevbest catalog-import awesomejev

Works forJev for Classification Jev for Ranking Jev for Tool calling

Used inJev for Software & DevTools

Languages: Python

Highlights

No dependencies beyond the standard library; runs on Linux, macOS or any python3.
Single ask call batches independent questions about the same state.
Skill file plus use-cases note document flags and answer reading.
Confidence reports how peaked a distribution is, not a prose hedge.

Watch out

It needs Python 3.9 or newer with only stdlib dependencies; credentials come from the environment or a JSON creds file and info never prints the key. Copy reviewed against the repository README on 2026-10-04, not from running it.

Questions about this entry

Short answers drawn from the linked project or page.

More Jev resources like this

jev-judge-mcp

MCP server over stdio that gives coding agents fourteen Jev judgment tools, turning evidence plus an enumerable question into auto, review, or escalate actions from calibrated probabilities.