Skip to main content

SESM Changelog

0.3.0 - 2026-06-11

  • Promoted the public-review candidate profile to v0.3.0 after the llm.interpretation_hints vocabulary change.
  • Added privacy considerations, conformance rules, submission pitch, contribution guidance, code of conduct, and license.
  • Added Validate-SESM-Safe.py as the safe-profile validator.
  • Added valid, invalid, and warning fixtures for safe-profile validation.
  • Expanded tests to cover safe-profile validator behavior.
  • Clarified schema compatibility for both 0.2.0 historical metadata and 0.3.0 public-review metadata.
  • Updated security support and schema wording for the 0.3.x candidate profile.

0.2.3 - 2026-06-11

  • Renamed the LLM hint field to llm.interpretation_hints before external review to reduce agent authority confusion.
  • Updated SESM schema, specification, README, and threat model language to frame LLM-facing fields as non-authoritative interpretation context.

0.2.2 - 2026-06-11

  • Added external-review documents: explainer, safe profile, threat model, security policy, validator rules, and reference implementation map.
  • Clarified that SESM metadata is untrusted input and does not make arbitrary SVG safe.
  • Registered the public review documents in the SESM suite manifest.

0.2.1 - 2026-06-10

  • Added SFDS two-layer suite metadata to the SESM standard manifest.
  • Added a SESM suite map example.
  • Clarified README, adoption, and validation language for suite conformance versus SVG metadata readiness.

0.2.0 - 2026-06-10

  • Added WGS/SFDS governance wrapper.
  • Preserved existing SESM spec, schema, tools, and tests as authoritative.