Proof path FAQ

Answer proof-path questions without losing the evidence boundary.

This FAQ is for readers who already know they need a proof path and want bounded answers about fields, tiers, coverage, limitations, gaps, review packets, and public-safe material.

Public claim level: concept. No public conclusion without evidence. The FAQ explains how to read deterministic static evidence trails; it does not create a new proof source or approval flow.

Placement

This FAQ answers repeated proof-path questions after topic selection.

/questions/Broad question-to-surface routing; this FAQ starts after the reader has chosen proof paths.
/proof-paths/Canonical proof-path overview and route family entry point; the FAQ uses question-and-answer explanations.
/proof-paths/tour/Guided step-by-step reading flow; this FAQ handles repeated objections and edge cases.
/evidence/Broader evidence vocabulary and artifact-shape explanation; this page applies those terms to proof-path questions.
/limitations/Canonical boundary surface; this page explains why a limitation travels inside a proof path.
/static-vs-runtime/Static evidence versus operational telemetry boundary; this page explains how static proof paths should be read.
/review-claim-checklist/ and /review-claim-checklist/drill/The checklist is the decision ritual; the drill gives practice rows and an answer key; this FAQ explains reading questions before that decision.

FAQ

Each answer keeps the claim, fields, limitation, and next owner attached.

What is a proof path?

A proof path is a public-safe route or reference trail from a public claim to supporting static evidence, rule ID or rule family, evidence tier, coverage label, limitation, and source context.

Stop if the trail cannot name the fields that carry the evidence.

How do I read a proof path?

Read in this order: claim, public claim level, proof path, rule ID or rule family, evidence tier, coverage label, commit or public-safe source context, extractor version or schema family, limitation, non-claim, and next owner.

If a required field is missing or private-only, stop, downgrade, hold, or route the question to the next owner.

What do evidence tiers mean?

Tier1Semantic is compiler-resolved symbol evidence, Tier2Structural is known framework or project structure evidence, Tier3SyntaxOrTextual is syntax-only or text evidence, and Tier4Unknown is an analysis gap.

A higher tier does not mean complete coverage. A lower tier can still be useful when the uncertainty stays visible.

What do coverage labels mean?

Coverage labels are visibility and confidence boundaries. They can be full, partial, reduced, unknown, unavailable, future-only, or gap-labeled states.

Preserve reduced or unknown labels instead of normalizing them upward.

Why do limitations matter?

A limitation is part of the claim boundary, not a footer. It must travel with any repeated claim because it says what the static evidence can and cannot support.

Do not repeat the claim after dropping the limitation, non-claim, or owner follow-up.

What do public claim levels mean?

A public claim level tells the reader how strong the public surface is allowed to be. This FAQ is concept-level guidance even when it links to demo-backed pages.

Do not upgrade concept wording into demo or main wording unless a recorded public-safe evidence surface supports that stronger claim.

What should I do when evidence is missing?

Stop, downgrade, label the gap, keep the claim internal, or hand off to the next owner. Missing evidence is not filled by confidence, seniority, repetition, or pressure.

The public answer is the gap or owner follow-up until the evidence-bearing field exists.

How do proof paths relate to review packets?

A review packet can gather proof paths, limitations, review notes, public claim levels, and owner follow-ups so a human can inspect what is known, partial, missing, or private-only.

The packet carries evidence and questions; it does not convert missing evidence into a stronger answer.

What should agents and reviewers preserve?

Human and automated readers must preserve the proof path, rule ID or rule family, evidence tier, coverage label, limitation, non-claim, and public claim level.

Do not repeat or summarize a claim after dropping those fields.

Static Evidence Non-Claims

What can static evidence not prove?

Public-Safe Material

Can a proof path use private or raw artifacts?

Public proof paths link only to public-safe summaries, checked-in docs, public routes, or sanctioned demo artifacts. Raw facts, raw SQLite, analyzer logs, raw source snippets, raw SQL, config values, secrets, local paths, raw remotes, generated scan directories, private sample names, raw command output, hidden validation details, and credential-like values are not public proof material.

Safe Answer Patterns

Use bounded verbs and keep the limitation in the sentence.

InspectThis proof path shows a static reference under Rule example.rule.family with reduced coverage; the limitation still applies. This is an illustrative pattern, not a real TraceMap finding.
Label the gapThe evidence is missing, so the public answer is a gap or owner follow-up, not a stronger conclusion.
Carry the packetThe review packet can carry this proof path and limitation to a reviewer; it does not approve the release.
Use bounded verbsinspect, follow, compare, check, record, downgrade, hold, label the gap, hand off, and escalate.

Unsafe Answer Patterns

Do not say or imply unsupported conclusions.