Deterministic check
A deterministic check is a repeatable validation procedure that applies explicit rules or conditions to determine whether an outcome is valid or meets specified criteria.
Spec-Driven Development
Deterministic verification is verification based on repeatable checks that produce evidence of whether a specified outcome has been achieved.
- How it is used
- Turns specified outcomes into evidence of alignment before implementation is considered complete.
- How deep it goes
- Works through a test-first failure-and-pass sequence and redirect assertions; leaves the choice of HTTP status to specification authoring.
- Examples
- Uses a short-link GET request with assertions for HTTP 302 and the destination Location header.
- What you can do after
- Practice/perform it by writing a check before implementation, confirming its initial failure, and rerunning it to verify the outcome.
Read how Spec-Driven Development explains deterministic check →
Vibe Engineering
A deterministic check is an automated validation whose output is predictably determined by its inputs and explicit rules.
- How it is used
- Uses deterministic checks as blocking gates while reserving model review for advisory judgments.
- How deep it goes
- Explains blocking authority, failure modes, and structured parsing, then demonstrates a pre-commit implementation.
- Examples
- Uses gitleaks, jq-parsed JSON verdicts, staged Git diffs, and advisory AI review in a pre-commit hook.
- What you can do after
- Apply a layered Git gate that blocks deterministic failures and reports model concerns without creating bypass incentives.
AI Agents in Action, Second Edition
A deterministic checkpoint is a programmatic or schema-based control point that evaluates a condition repeatably without relying on stochastic model judgment.
- How it is used
- Controls agent flows and structured-output evaluation with repeatable pass/fail checks.
- How deep it goes
- Defines code or schema checks and contrasts exact tests with fuzzy and human evaluation, but does not specify implementation code.
- Examples
- Uses schema validation, expected outputs, correct tool calls, and known-good answers as checkable cases.
- What you can do after
- Apply it via code or schemas to route flows and test exact agent behavior.
Read how AI Agents in Action, Second Edition explains deterministic check →
Also covered in
- AI Model EvaluationIn early access (MEAP)
What to know first
- evaluationEvaluation is the systematic assessment of a system, model, process, or outcome against relevant criteria or objectives using evidence.
- heuristicA heuristic is a practical rule or approximate method for guiding judgment, search, or problem solving without guaranteeing an exact or optimal result.
- human-in-the-loopHuman-in-the-loop is a system or workflow design in which people participate in reviewing, guiding, correcting, or authorizing an automated process.
- language modelA language model is a computational model that represents patterns in linguistic sequences to assign probabilities, predict, generate, or evaluate language.
Related concepts
- LLM-as-a-judgealternative to · LLM-as-a-judge is an evaluation technique in which a language model assesses or compares outputs according to specified criteria.
- implementation gapaddressed by · An implementation gap is a discrepancy between behavior intended by a specification and behavior or coverage in an implementation.
- security boundaryused by · A security boundary is a separation between components or contexts across which access, trust, or information flow is controlled.
- evaluationused by · Evaluation is the systematic assessment of a system, model, process, or outcome against relevant criteria or objectives using evidence.
Questions about deterministic check
What is deterministic check?
A deterministic check is a repeatable validation procedure that applies explicit rules or conditions to determine whether an outcome is valid or meets specified criteria.
Which Manning books explain deterministic check?
Spec-Driven Development, Vibe Engineering and AI Agents in Action, Second Edition.
What should I know before learning deterministic check?
Manning books that teach deterministic check usually introduce evaluation, heuristic, human-in-the-loop and language model first.
What is the difference between deterministic check and LLM-as-a-judge?
A deterministic check is a repeatable validation procedure that applies explicit rules or conditions to determine whether an outcome is valid or meets specified criteria. LLM-as-a-judge is an evaluation technique in which a language model assesses or compares outputs according to specified criteria.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.


