Security & claims

This section is the honest answer to "what is Myelin actually proving?" It's deliberately precise: every claim Myelin makes has a corresponding evidence path, and every evidence path has a boundary where it stops being a claim.

Pages

• Claim ladder

  ---

  The four-tier ladder from "designed to stay close to CKB semantics" up to "CKB-aligned adjudication path." Each tier has a specific report that establishes it.

• Evidence paths

  ---

  The four evidence paths (execution, projection, court, settlement), what each one proves, what each one doesn't.

• Threat model

  ---

  What Myelin protects against today, what it explicitly does not protect against, and what's outside scope.

Why this section exists

Most protocol docs bury the security boundary in marketing copy. Myelin's project-level docs (MYELIN_USE_CASE_POSITIONING.md, MYELIN_PRODUCTION_GATE.md, etc.) make a point of saying exactly what the current evidence is and isn't. This section is the user-facing version of the same discipline.

Read this section before you trust any number in the rest of the docs.