Jev instruction file side effect

Review an instruction file for unexpected side effects

Give the instruction excerpt and its disclosed purpose. Compare requested data transfer, hidden execution or unrelated changes with behavior that serves the stated task.

Which unexpected effect is explicit?Live API
Try an example

The transfer is unrelated to the disclosed formatter task.

Source: API-collected workflow inspiration; original demonstration below

Source and example scope

The API-collected roundup links the cited project, whose pinned code provides the decision purpose. This demonstration separately authors its Choice rubric, priorities, missing-context outcomes and synthetic inputs. It neither runs the upstream application nor verifies its performance, security or execution claims.

Compare excerpt with declared purpose.

Decision rules and labels
Data sendingHidden executionRule overrideUnrelated changePurpose alignedNeed purpose
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

An instruction file is supposed to instruct; the issue is what it asks for.

The formatter preset asks to read selected Markdown files and show the resulting diff. Those instructions match its disclosed purpose. An undisclosed conversation upload adds a separate effect the stated task does not require.

Global configuration changes and permission overrides deserve their own observations. The page labels the behavior explicitly present in a short excerpt, rather than assuming that any imperative sentence is an attack.

Purpose aligned is limited to the supplied excerpt. It is not a safety certification for a package, its dependencies or code that was not provided. The window never installs the file or runs its commands.

The API-collected roundup links the cited project, whose pinned code provides the decision purpose. This demonstration separately authors its Choice rubric, priorities, missing-context outcomes and synthetic inputs. It neither runs the upstream application nor verifies its performance, security or execution claims.

Illustration of reviewing messages against supplied policy categories
Illustration of the decision flow.
How it works

Try the example in four steps

Use the sample to understand the rule, then test your own input for instruction-file side effects.

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

    Add the information to check

    Give the instruction excerpt and its disclosed purpose.

  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 Data sending, Hidden execution, Rule override, Unrelated change, Purpose aligned. 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 Need purpose before using a result. The label is a decision about the supplied information, not an executed action.

Our API test

Recorded API observations

Recorded API observations
Original inputObserved labelLeading probability
Undisclosed exportData sending100%
Concealed commandHidden execution100%
Disable a guardRule override100%
Global preference editUnrelated change100%
Disclosed formatterPurpose aligned99.0%
Purpose absentNeed purpose98.0%

These are six original synthetic inputs, tested once per request revision. The percentage is the leading option probability, not measured accuracy. Authored expectations and any disagreements remain in the downloadable evidence. 6 of 6 results matched the authored fixture expectations.Verified

Understand the results

Replay the saved synthetic requests to inspect each recorded decision

Replay the saved synthetic requests to inspect each recorded decision. Authored expectations and model disagreements remain available; these examples are not a production accuracy benchmark.

Using the result in your application

The single-request SDK runner uses your private TYPESAFE_API_KEY. The offline consumer reads saved observations, checks request hashes and prints a pending output. It does not call the model or change any external system.

Runnable offline example

Run one sample, then inspect the local packet

View the code example
Single-request replay
npm install @typesafe-ai/sdk@0.6.0
# Set TYPESAFE_API_KEY privately in your shell.
node run-choice-sample.mjs samples.json case-1
python preview_choice_workflow.py samples.json saved-api-results.json
Scope of this demo

Where this decision fits

Useful scope

  • Inspecting an agent rule or skill excerpt
  • Comparing disclosed purpose with requested side effects
  • Preparing specific concerns for a human review

Outside this demonstration

  • Certifying an entire package as safe
  • Executing downloaded code
  • Inferring hidden behavior that is not supplied
Example FAQ

Questions about instruction file side effect.

Input scope and downstream handling determine what this judgment means.

Why are ordinary instructions allowed?

A rule or skill file exists to guide an agent. The comparison concerns undisclosed or unrelated behavior, not the presence of instructions.

Does Purpose aligned approve installation?

No. It describes the provided excerpt and purpose. Installation requires a separate review of the actual package and its effects.

Is the leading probability an accuracy measurement?

No. It is one value in the returned distribution over this request’s choices. The six synthetic fixtures have authored expectations; they do not measure production accuracy.

What happens on an API failure?

The window displays an error without a fresh judgment. An incomplete-input outcome describes the supplied evidence and is not used as a fallback for provider failure.

Try your own instruction file side effect input

Edit the input above, or open the Playground to explore another decision.