Jev in production › Agent guardrails and approvals

DSH Jev Guard

DeepSeek Harness plugin on tools/pre-execute asks Jev whether a shell command will irreversibly delete or overwrite real data, after offline hard rules, and maps the probability to allow, revise or block; 90.4% on a 114-case calibration set (author).

Open on GitHub ↗

90.4%measured, as published by the source
Use
Agent guardrails and approvals
Industry
Security
Form
Open-source tool
Stage
In production
Plugs into
DeepSeek Harness
Listed
2026-09-22
Found via
discord
Repository
7starsseeker/dsh-jev-guard
Stars
0
Forks
0
Last push
2026-09-29
Language
JavaScript
License
MIT

The README opens with

A pre-execution safety valve for DeepSeek Harness (DSH): before a command actually runs, it asks one question — "will this irreversibly delete or overwrite your real data?"

It judges with TypeSafe Jev — a "System One" model that produces no prose, only a structured decision. Yes/no questions are what this kind of judgment measures most accurately (114-case calibration, 90.4% with the Chinese text sent as-is). It hooks into DSH's native tools/pre-execute, so it is a hard interception, not a "please be careful" nudge to the model.

Badge

For the project's own README, linking back here:

Listed in Jev in production

Also used for agent guardrails and approvals