Jev fictional canon claim check

Check a fictional claim against supplied canon

Give the claim and reference passage. Compare facts established by the fictional canon with contradictions or plausible additions the passage never supports.

Check every part of the claim against the reference.Live API
Try an example

Original synthetic input. Edit the task or evidence to make a new API request.

Source: Workflow inspiration on X

Source and example scope

The inputs, rubric and preview are original. This window demonstrates the stated decision only; it does not reproduce the source application or its measurements.

Does the entire fictional claim follow from the supplied canon?

Decision rules and labels
SupportedContradictedNot establishedMissing canon
A real response from the Jev API

Edit any input to explore a different decision. No request is sent until you run it.

Your next decision

Not run yet

Choose an example or write your own input, then run it to see the ranked choices.

Results are not stored by this site. Your submitted text is processed by the API provider.

The decision

How does Jev fictional canon claim check use the supplied evidence?

Jev fictional canon claim check compares one complete dispatch claim with a reference passage. Our original canon says that a fictional council opened negotiations. A claim that negotiations began can be supported; a claim that an agreement was signed adds an outcome the passage never established. The model returns a label and writes no story.

An API-collected X post described a fictional-canon lab using Jev to assess dispatch claims. Our exercise adapts that idea to an explicit four-outcome Choice and fixed local reports. It does not run the author's scoring pipeline, reproduce their fictional timelines or verify real historical events. Every canon and claim below is original synthetic text.

The entire claim matters. A sentence can begin with a supported fact and end with an unsupported addition. The mixed sample preserves that problem rather than giving the whole sentence credit for its first half. Missing canon is also separate from a supplied canon that simply leaves the outcome unknown.

Original Jev fictional canon claim check workflow diagram showing supplied evidence, a typed decision and the local preview.
Original workflow illustration. The output is a local preview, not an executed action or source screenshot.
How it works

Try the example in four steps

Use the sample to understand the rule, then test your own input for fictional canon evidence.

  1. Illustration of adding the text or evidence for a decision

    Add the information to check

    Give the claim and reference passage.

  2. Illustration of defining a question and its allowed choices

    Confirm the question and labels

    Open the decision rules and keep the intended distinction between Supported, Contradicted, Not established, Missing canon. Edit the choices only when testing your own rule.

  3. Illustration of sending one input and comparing the returned choices

    Run the decision

    Use a sample first, then try your own input. Compare the selected label with the other returned probabilities.

  4. Illustration of reviewing a result and an uncertain item before using it

    Review the next step

    Review Missing canon before using a result. The label is a decision about the supplied information, not an executed action.

Our API test

What the six original inputs returned.

What the six original inputs returned.
Original test inputObserved labelLeading probability
Established dispatch detailSupported100%
Contradict the named dayContradicted100%
Negotiations are not an agreementNot established62%
One supported part plus an additionNot established73%
No reference passageMissing canon100%
Supported wording variationSupported95%
  • Measured adapter round trips: 474-1289 ms. Estimated input cost: $0.00002297-$0.00002335 per request. These observations are not a latency guarantee or invoice.

6 actual Jev API calls on October 1, 2026. Original synthetic inputs check this rubric and handoff; they do not establish held-out accuracy. Probabilities are model outputs.Verified

Understand the results

The matching-negotiation sample repeats the complete fact established by its passage

The matching-negotiation sample repeats the complete fact established by its passage.

Read the example notes

Supported means supported by that supplied fictional canon. It does not mean independently true in the real world. The offline payload retains realWorldVerified false, making that scope visible even when the model's leading probability is high.

The day-contradiction sample explicitly states that the fictional event was Monday and not Tuesday

The day-contradiction sample explicitly states that the fictional event was Monday and not Tuesday.

Read the example notes

That evidence supports Contradicted for a Tuesday claim. An omitted day would be a different problem. The rubric needs an explicit conflict before using this label, rather than interpreting every absent detail as false.

The signing sample illustrates missing support

The signing sample illustrates missing support.

Read the example notes

Opening negotiations does not establish a signed agreement. The canon gives no outcome, so Not established preserves the additional fact as unverified. This is a text-evidence exercise; it does not rely on assumptions about real diplomacy, customary timelines or what usually happens after a meeting.

Using the result in your application

The mixed claim combines the supported Monday opening with an unsupported signing.

Read the example notes

The question covers every asserted part, and the original expectation remains Not established. A production evaluator should split longer dispatches into clearly identified claims before making these checks. Our consumer receives one claim in each sample and performs no sentence splitting or source retrieval.

The missing-canon case has no reference passage to inspect

The missing-canon case has no reference passage to inspect.

Read the example notes

It cannot become supported merely because the claim sounds ordinary. The paraphrase case provides a contrasting positive example: responsible for the archive and oversees the archive describe the same supplied responsibility. Use more demanding paraphrases and partial contradictions in an independent evaluation.

Source and example scope

A real evidence workflow needs reliable reference provenance, passage coverage and current revisions.

Read the example notes

A model cannot detect an omitted source fact that it never receives or independently authenticate a fabricated canon. Preserve the exact passage with the full response. The downloaded replay returns local scoped verdicts only; it does not revise, publish or establish the truth of a dispatch.

Using the result in your application

Download the exact sample requests and saved observations below.

Read the example notes

Expected labels are original fixture checks and stay outside the request sent to Jev. The observation file preserves the complete probabilities and the hash of each exact request. Editing a task, rubric or candidate requires a fresh response; a matching sample name does not make an older answer current.

Using the result in your application

The JavaScript runner makes one independent SDK request for a chosen sample ID.

Read the example notes

It reads TYPESAFE_API_KEY privately from the environment and has no automatic retry. The Python consumer operates offline: it checks the rubric, labels, unique sample identities, request hashes and valid probability distributions before selecting a fixed local preview. A tied leading result remains pending for review.

Source and example scope

The page reports measured adapter round trips and estimated input cost for the recorded calls.

Read the example notes

Those values describe these requests, not a provider latency guarantee or a billing statement. Six deliberately authored fixtures can check the stated behavior, but they do not establish accuracy on an unseen dataset. Keep disagreements and missing-input examples when testing an integration.

Runnable offline example

Replay one request and inspect the saved preview.

View the code example
Replay commands
npm install @typesafe-ai/sdk@0.6.0
# Set TYPESAFE_API_KEY privately in your shell.
node run-choice-sample.mjs samples.json support
python preview_choice_workflow_v2.py samples.json saved-api-results.json
Scope of this demo

Where does this decision fit?

Useful with explicit supplied evidence

  • Check whether an entire original fictional dispatch claim is supported, contradicted or not established by the supplied canon. Preserve an absent canon separately. The model does not write a dispatch, verify real history or recover unread sources.
  • Inspecting the original task and a fixed output before integrating external tools.
  • Comparing an explicit expected label with the complete recorded distribution.

Needs separate implementation or evidence

  • Inferring facts absent from the supplied input.
  • Treating a semantic selection as authorization to act.
  • Executing an external action or generalizing six fixtures into an accuracy claim.
Example FAQ

Questions about this original decision workflow.

Keep relevance, compliance and execution separate.

Does Supported mean the claim is true?

It means the entire claim is established by the supplied fictional passage under this rule. RealWorldVerified stays false; no outside source is checked.

Why are missing support and contradiction different?

An absent fact does not establish the opposite. Contradicted requires an explicit conflict; Not established preserves a claim the supplied passage does not support.

What happens to a partly supported claim?

Every asserted part must be supported. A sentence with an added unestablished fact remains Not established under this contract.

Does this split or rewrite a dispatch?

No. Each sample supplies one claim. Text splitting, writing and editing belong to separate application components.

Is this the source's Score pipeline?

No. It is an original bounded Choice exercise with four stated outcomes. It does not reproduce source thresholds, escalation, timelines or performance results.

Try a task from your own workflow

Change the supplied task or evidence here, then inspect the live response. Playground supports another explicit decision.