---
ref: tool:typesafe/answer-typed-questions
name: Answer typed questions
company: company:typesafe
workflows: [workflow:ad-persona-panel, workflow:ai-answer-positions, workflow:cancellation-winback, workflow:chat-buyers-to-sales, workflow:cold-email-reply-triage, workflow:community-buyer-threads, workflow:competitor-ad-teardown, workflow:customer-call-risk, workflow:draft-style-check, workflow:feature-requests-to-linear, workflow:hiring-signal-outbound, workflow:icp-account-qualification, workflow:inbound-lead-routing, workflow:job-title-personas, workflow:linkedin-network-buyers, workflow:linkedin-opener-check, workflow:linkedin-reply-triage, workflow:promoter-case-studies, workflow:referral-reply-leads, workflow:shared-inbox-triage, workflow:signup-fit-scoring, workflow:support-expansion-signals, workflow:support-ticket-churn-risk, workflow:visitor-company-fit]
access: [api]
tags: [capability:classify-signals, category:ai-model, has:api]
docs: https://docs.typesafe.ai/api
updated: 2026-09-29
---

# Answer typed questions

Answers named choice, score and yes-or-no questions about one text or JSON state with Jev in one pass, with each answer's probabilities and a confidence for choices and scores.

Note: Send `state`, `model` (`jev-latest`) and named `questions`, each with `type`, `instructions` and `criteria`: a choice maps labels to descriptions, a score lists its levels in order, a noul's is optional. Read a score by its most likely level; send unsure answers to a person.

## Set up

Use the first option your agent supports.

### API (official)

- Base URL: https://api.typesafe.ai
- Endpoint: `POST /v1/systemone`
- Auth: send the header `Authorization: Bearer $TYPESAFE_API_KEY`
- Get a key: https://console.typesafe.ai
- Docs: https://docs.typesafe.ai/api

Note: Put every question about one record in one request. `GET /v1/models` lists the models and is the cheapest check of a key; a request over the rate limit gets a 429 with a Retry-After header.

Before anything else, confirm access with the cheapest read-only call, like a list or a search. Never send or change anything to test access.

## Rules

- Ask the user before anything that sends messages, costs money, or changes data, and say how many records it touches. One approval covers a batch the user has seen.
- Never print API keys.
