[Docs] Create VISION.md and vision artifacts for Sistent #1830

Objective

Run the /vision skill in layer5io/sistent and produce its deliverable: a VISION.md at the repository root, following the skill's structure rather than a hand-rolled vision document.

Scope and constraints:

• Documentation only. Do NOT change build, application, or extension behavior.
• VISION.md at the repo root only; do not add one to any subdirectory or other repository.

Quality bar:

• Nothing generic. Anything that could be copy-pasted onto another project is a defect.
• Every substantive claim traceable to this repo, the downstream projects (e.g. meshery/meshery), or published material on https://layer5.io/projects/sistent. No asserted roadmap, adoption, or capability that cannot be pointed at. Claims describing direction rather than shipped state are deliberately marked as such in the evidence sheet.
• Match the repo's established voice and Layer5/Meshery documentation conventions.
• No em dashes (project-wide edict); plain dashes only. This is deliberate, not an oversight.
• No AI attribution anywhere.

Deliverables, all intended and to be preserved:

• VISION.md at repo root.
• docs/vision/vision-evidence.md - claim-by-claim evidence sheet; every line of VISION.md traces to it.
• docs/vision/vision-hypotheticals.md - ten stress-test hypotheticals with the author's verdicts and reasoning quoted VERBATIM, plus a changelog mapping each verdict to the edit it produced. Durable calibration material that belongs next to the vision. The verbatim block quotes are required; do not paraphrase or condense them.
• The interactive review board (docs/vision/review-board.html, docs/vision/review.css) was the review instrument and was DELIBERATELY DELETED once all ten cards were answered; its content is fully superseded by vision-hypotheticals.md. Its removal is intended. Do not restore it or flag it as an accidental deletion.