Moderation input contracts with Jev
UGC needs a category, a severity, and an allow/review/remove decision. Jev scores the text you provide against your policy excerpt. Code enforces.
This unofficial page is the input contracts slice of the content moderation pack. Intent: apply the Jev (TypeSafe System One) decision model to content moderation input contracts. Primary search language: Moderation Jev input contracts. Confirm patterns on docs.typesafe.ai. This site does not sell, issue, or proxy TypeSafe keys. Use a credential you already have from the console or a documented gateway.
Independent angle (cover ≠ clone): Policy-as-criteria + confidence abort + human pack — not a clone of a moderation-API landing page or rival recipe IA.
Moderation use-case context
An input contract is the allow-list of fields you will ever POST for content moderation. It is a decision contract for the user-generated post or message: if a field is not named in instructions, it should not be in state. That is how you beat noisy “dump the object” integrations — the rival-intent failure mode — without cloning anyone’s IA.
Hub: Use cases. Compare, when the other tool is the real job: moderation APIs.
Input Contracts inputs
Documented System One inputs: state (string, object, or array of text) and a questions map. English is the primary training language. Images, audio, and video are not accepted.
Allow for content moderation:
{
"post": { "id": "p-209", "text": "…", "locale": "en" },
"policy": { "hate": "…", "spam": "…", "illegal": "…" },
"author": { "strikes": 1, "age_gate": "18+" }
}Bind paths: post.text, policy.hate, policy.spam.
Refuse at the wrapper (do not send):
- image/video bits (not accepted — run a vision stack elsewhere and pass text labels)
- the author’s entire post history as one blob
- secret moderator Slack
TypeSafe’s published list price for jev-1.13 is $0.042 per million input tokens (vendor claim — confirm on the models page); output tokens are free on that same page. Unused distractors still bill as input.
Decision signals and actions
The contract exists so each primitive stays atomic:
| Id | Type | Job |
|---|---|---|
category |
Choice | ok / spam / hate / harassment / illegal / other |
severity |
Score | nuisance → severe harm |
allow |
Noul | Would a trained moderator leave this up given policy.*? |
If a new CRM field appears, either add a question that names it or drop it. Do not “just include it.” Do not treat a Noul of 0.5 as a “medium” content moderation score — it means yes and no are equally likely. Conjunctions stay in your code.
Guardrails and escalation
Contracts are a guardrail: missing required text → do not call Jev (or ask a Noul “is enough information present?”). That is cheaper than a confident wrong category. TypeSafe’s confidence-gated examples use a lower bar for recoverable reads than for irreversible actions. Those numbers are illustrations. For content moderation, treat remove_or_ban as the high bar (removing content or issuing a ban). Tune on labels — see offline evaluation.
Evaluation and rollout notes
Version the contract (field list + criteria git SHA) next to the pinned model. Replay keep / review / remove gold from trained mods, plus category gold when either changes. Pin jev-1.13.0 (the versioned id) after you fit thresholds. jev-latest and the marketing line jev-1.13 can move. Log the response model. TypeSafe’s published list price for jev-1.13 is $0.042 per million input tokens (vendor claim — confirm on the models page); output tokens are free on that same page. Unused distractors still bill as input.
Official Python and JavaScript SDKs read TYPESAFE_API_KEY and retry documented 429/529. This site does not sell, issue, or proxy TypeSafe keys. Use a credential you already have from the console or a documented gateway.
Pack map
| Slice | Page |
|---|---|
| Graph and primitives | decision workflow |
What may enter state |
you are here |
| What to gather first | evidence collection |
| Atomic rules | policy checks |
| Act / review / abstain | confidence thresholds |
| Reviewer payload | human handoff |
| What to persist | audit trail |
| How it breaks | failure modes |
| Labeled replay | evaluation |
| Shadow → canary | production rollout |
FAQ
What happens if I send the whole warehouse row?
jev-1.13 loses accuracy as distractors grow (official jaggedness note). Drop image/video bits (not accepted — run a vision stack elsewhere and pass text labels). TypeSafe’s published list price for jev-1.13 is $0.042 per million input tokens (vendor claim — confirm on the models page); output tokens are free on that same page. Unused distractors still bill as input.
Can I send images of the artifact? No. State is text (string, object, or array of text). Transcribe first.
Where is the rest of the Moderation pack? Start with Moderation decision workflow and Moderation evidence collection. Cluster hub: Use cases.
Should we replace our moderation vendor with Jev? Only after a labeled bake-off you run. This page does not publish one. See Jev vs moderation APIs.
Can Jev moderate images? Not directly. State is text. Run a vision system, put labels/transcripts in state, then ask typed questions.
What this page does not claim
- Not a trust-and-safety certification.
- No published precision/recall.
- Not official TypeSafe.
- Official TypeSafe status, or that jev.pro issues API keys.
- That a schema-constrained answer is automatically factually correct.
Disclaimer
This is an independent unofficial site and is not affiliated with TypeSafe AI; official documentation is available at https://docs.typesafe.ai.
Primary documentation: https://docs.typesafe.ai. Hub: Use cases.
Sources
Public TypeSafe or adjacent documentation only. No private claims.