User repository on GitHub
Skill for Hermes, and other agents, to ask typesafe's jev
- Forks
- 1
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Tools & CLIs
Askjev — shantanugoel
A Hermes skill that calls Jev as a typed tiebreaker: the agent submits a state and a question, and Choice, Score, or Noul answers with probabilities let it act or escalate.
- Stars
- 41
- Language
- Python
- License
- MIT
- Reading time
- 3 min
- Published by
- Community
- Author
- shantanugoel
- Last verified
- Sep 2026
About this project
Hermes skill: call TypeSafe Jev (jev-latest) as a typed tiebreaker. Jev is not a chat model. It returns Choice / Score / Noul answers with probabilities so an agent can act or escalate. Put TYPESAFEAPIKEY in $HERMESHOME/.env. Never commit the key. Set ASKJEVDISABLED=1 in that .env, or delete the skill directory. See SKILL.md for when to call Jev, confidence gates, and pitfalls.
Quoted from the project's README. Read the full README.
Highlights
- Installs as a SKILL.md plus scripts under $HERMES_HOME/skills/autonomous-ai-agents/askjev.
- The README frames Jev as a typed tiebreaker rather than a chat model, returning Choice, Score, or Noul answers with probabilities.
- The CLI takes a state, a question, and named options: python3 scripts/askjev.py choice --state '...' --question 'Which path?' --option a='...'.
- TYPESAFE_API_KEY goes in $HERMES_HOME/.env, and ASKJEV_DISABLED=1 disables the skill without deleting it.
- scripts/eval.py runs a live reliability suite that needs the API key.
Quickstart
bash
mkdir -p "${HERMES_HOME:-$HOME/.hermes}/skills/autonomous-ai-agents/askjev"
cp -R SKILL.md scripts "${HERMES_HOME:-$HOME/.hermes}/skills/autonomous-ai-agents/askjev/"Watch out
MIT-licensed. Needs Python 3, a Hermes agent installation, and a TYPESAFE_API_KEY placed in $HERMES_HOME/.env.
Questions about this entry
Short answers drawn from the linked project or page.
A thin skill, and the README presents it that way. It installs as a SKILL.md plus scripts under the Hermes home directory and calls Jev as a typed tiebreaker rather than a chat model, returning Choice, Score or Noul answers with probabilities so the agent can act or escalate. The README documents no prompt engineering, retry or routing logic of its own, and defers the usage guidance to SKILL.md.
Set the disable flag to 1 in the same env file that holds your API key, or delete the skill directory. The key goes in the Hermes home env file and the README says never to commit it.
Only loosely. There is a live reliability suite, the eval script under scripts, that needs the API key, and SKILL.md covers when to call Jev, confidence gates and pitfalls. The README publishes no expected pass rate, no thresholds and no accuracy figures, so anything beyond the existence of that suite is not documented.
By hand. There is no package manager step: you create the skill directory under your Hermes home and copy SKILL.md and scripts into it, then put the API key in the Hermes env file. It needs Python 3 and an existing Hermes agent installation. The CLI takes a state, a question and named options, for example a choice call with two options.
MIT. Runtime requirements are Python 3, an existing Hermes agent installation, and a TYPESAFE_API_KEY placed in the .env file inside your Hermes home, which the README says never to commit. There is no build step and no dependency list beyond the standard library.
One command per judgment, run from the installed skill directory: python3 scripts/askjev.py choice --state '...' --question 'Which path?' --option a='...' --option b='...'. The state is the material to judge, the question is what you want to know, and each option is a named alternative with its own description.
Jev at jev-latest, framed explicitly as a typed tiebreaker rather than a chat model. It returns Choice, Score or Noul answers with probabilities so the calling agent can act or escalate, and SKILL.md carries the guidance on when to call it at all, how to gate on confidence, and what the pitfalls are.
Other listings of this project
Forks and copies of this project in the directory. This entry is the canonical one.
Claude-driven browser automation where plain-English chat drives Brave or Chrome, Jev makes on-page decisions, and a guard freezes irreversible clicks.
Jev Demos & Experiments#community#jevbest#catalog-import
Community
ranjan2829/AskJevTypeScript
ranjan2829/AskJevTypeScriptAn unofficial MCP server exposing one ask tool: your agent hands Jev some material and a list of plain questions, and gets typed answers with probabilities back in one round trip.
Jev Agents & MCP#community#jevbest#catalog-import
Community
pZacca/askjevTypeScript
pZacca/askjevTypeScriptMore Jev resources like this
A streaming grep-by-meaning CLI that judges each line with one Jev yes/no question and prints the lines above a threshold, built to sit behind tail -f, kubectl logs -f, or journalctl -f.
Jev Tools & CLIs#community#python#cli
Community
allebee/jevgrepPython
allebee/jevgrepPythonA collection of Jev demos, workflows, and agent skills for coding agents: five installable skills and 108 scenarios, installed by pointing Codex, Claude Code, or OpenCode at an agent prompt.
Jev Tools & CLIs#community#python#skills
Community
wuyoscar/jev-skillPython
wuyoscar/jev-skillPythonA collection of 26 production-ready agent skills for Claude Code, Cursor, Kiro, Windsurf, and OpenCode; four skills call TypeSafe Jev for calibrated Score and Noul judgments and fall back to heuristics when it is unavailable.
Jev Tools & CLIs#community#python#skills
Community
fabricioctelles/skillsPython
fabricioctelles/skillsPythonClaude Code, Codex, and pi plugin plus MCP server and npm library that hands steps needing no text output to Jev: one batched call answers typed questions about a state, and an opt-in PreToolUse gate can only deny or ask.
Jev Tools & CLIs#community#javascript#claude-code
Community
shitianfang/jev-useJavaScript
shitianfang/jev-useJavaScriptA semantic code search CLI that asks one yes/no Jev question per function and ranks the returned probabilities, screening a whole codebase for a pattern in seconds for cents.
Jev Tools & CLIs#community#python#cli
Community
sufianetaouil/everyPython
sufianetaouil/everyPythonClaude Code hook that scores installed skills against your prompt with Jev and writes skillOverrides so irrelevant skills stay out of context; a local TF-IDF ranker runs when no API key is set.
Jev Tools & CLIs#community#javascript#claude-code
From the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.