claudemods

Modsjev-guardrails

jev-guardrails 0.1.0 Immutable snapshotLatest

Pinned to commit e945056ea1826a54351f59d85ab95b41535d8d6c, captured .

Download 0.1.0 source archive Choose something to compare with View commit on GitHub

More actions and subscriptions

All versions

Current package description (not historical)

Screens every prompt going into the conversation and every reply coming out of it with TypeSafe's Jev, a System One decision model, the way TypeSafe's guardrails cookbook does: one request per message asks a battery of hazard questions (jailbreak, harm or a crime, a diagnosis or a dosage, self-harm) and scores how much harm complying would do. The thresholds live here, in a named policy (strict or permissive), and turn the probabilities into pass, review (the user is asked), block (the prompt is dropped, or the reply withheld) or support. Falls back to the engine's built-in classifier when no key is set.

Install

No verified marketplace install command.

This plugin is not listed in a verified marketplace. The registry has no installer of its own; download the source archive or read the install notes in the upstream repository.

Contents 18

Config18
Dependencies, compatibility and skill declarations

Recorded dependencies 0

No dependency requirements were recorded from plugin.json or marketplace entries. Only manifest fields saved with the snapshot are shown here; dependency declarations in other files are outside the current collection scope.

Recorded compatibility

No compatibility requirement was recorded for this snapshot from plugin.json or a marketplace entry. Only manifest fields saved with the snapshot are shown here. Declarations in a skill's SKILL.md header are collected separately and, when present, shown on their own under “Skill file declarations”; they never overwrite or fill in this value.

Versions of this package1
VersionCommitCapturedComponentsActions
0.1.0 Latest e945056 18 View Download Compare…