ai // LLM EngineeringSTUDY

LLM Review — Bounded Context + Provenance Prompt

Constrain an AI review to supplied context and require explicit separation between documented facts, observations and inference.

llmcontextprovenancereview

Context Pack →Open Workspace →

Trust & provenanceSTUDY

Evidence tierSTUDYOfficial sources0Other sources0Last verified2026-10-05

Evidence tier describes my hands-on relationship with the content. Official sources describe platform contracts/documentation. Production observations should not automatically be treated as official limits.

EvidenceSTUDY

Difficultyfoundation

Verified2026-10-05

Use when

Using AI to review technical artifacts where unsupported assumptions would be costly.

Use only the context below.

Rules:
1. Separate documented facts, production observations, inference and unknowns.
2. Do not invent fields, limits, dependencies or business rules.
3. When evidence conflicts, show the conflict instead of reconciling silently.
4. Ask for missing context when a decision depends on information not supplied.

Why it works

The prompt makes epistemic boundaries part of the requested output rather than relying on model restraint implicitly.

Operational notes

A bounded prompt reduces risk; it does not replace deterministic validation.

Prefer machine-readable evidence when the result will feed automation.

LLM Review — Bounded Context + Provenance Prompt

Keep following the context.