Skip to main content

DDR-001 — Bracket-S monogram as the SecurityV0 mark

Status: Accepted Date: 2026-05-05 Supersedes: —

Context​

SecurityV0 needed a primary mark for the platform UI, login pages, favicon, marketing, and partner surfaces. Three rounds of design exploration produced eight candidate marks across two trademark-research scans. By 2026-05-05 the question was: pick one and lock the geometry.

The mark needs to:

  1. Read as a security platform, not a generic SaaS chip or an AI tool.
  2. Carry the product's deterministic / engineering register (path IDs, scanner output, no probabilistic scoring).
  3. Survive a non-trivial trademark constraint — V0 overlaps Vercel's v0 AI-codegen wordmark.
  4. Scale cleanly from 16 px favicon to 128 px launcher without redrawing.
  5. Pair with a two-tone wordmark (Security + demoted V0) without competing for attention.

Decision​

The mark is [S] — an uppercase S in JetBrains Mono 700, set between two SVG-path brackets, on a filled tile.

The geometry is locked at a 32-px viewBox with all other dimensions derived from M (mark side length). Full spec: 01 · Logo system.

The two-tone wordmark (Security + demoted V0) ships with the mark but is structurally subordinate. The [S] is the mark; the wordmark is a typographic companion.

Alternatives considered​

A · Bracket monogram [S] — chosen​

The brackets visually encode the engineering register. Code-style, version-cue, no AI-tool feel.

  • Trade-off: bracket monograms are common in dev-tool branding (Bitbucket-era, dozens of CLI tools). Silhouette is generic.
  • Why we accepted: the engineering register matters more than silhouette uniqueness for this product. Differentiation comes from voice, color discipline, and the surface contract — not the silhouette.

B · Layered shield — rejected​

Two offset shields ("defense in depth"). Visually encodes the security category.

  • Rejected: shields are saturated in security branding (McAfee, Norton, hundreds of MSSPs). Not a legal conflict, but a memorability failure — the mark blends into category noise.

C · Orbit / V0 dot — rejected​

A circle with a tracking dot, V0 inside.

  • Rejected: orbit + tracking-dot motifs are heavily used by AI/agent products (OpenAI, Anthropic Atlas-era, Cursor adjacent). Combined with V0, the impression skews "AI tool" when this is a security platform. Medium trademark risk.

D · Wordmark only — rejected as primary, kept as fallback​

A typographic-only mark. Lowest collision risk.

  • Rejected as primary: no recall surface at favicon size. The mark IS the wordmark, which forces every consumer to render text legibly at 16 px — fragile.
  • Kept as a fallback for body copy and footers where the full lockup is overkill.

E · Stacked block — rejected, held​

Two-layer block referencing the "posture / evidence" stack from the product itself.

  • Rejected (this round): the conceptual link to the product is strong but requires explanation. The bracket monogram works without commentary.
  • Held: re-evaluate if the brand voice ever leans heavily into the "evidence stack" framing.

F · Notch / chevron — dropped (high TM risk)​

A geometric V cut into a square.

  • Rejected: too close to Vercel's triangle — same negative-space-triangle motif Vercel has owned since 2015. At small sizes the resemblance is strong.

[V0] uppercase / [v0] lowercase — rejected (high TM risk)​

Replacing the S inside the brackets with V0 or v0.

  • Rejected: putting v0 inside its own bracketed badge is precisely Vercel's wordmark format. Medium-to-high trademark risk; explicitly warned in the v3 trademark scan.
  • The [S] keeps V0 as a typographic suffix only, which is the safer hedge.

Consequences​

What this decision constrains:

  1. The mark is [S]. Forever, until a new DDR supersedes this one. No [V0], no [v0], no shield, no orbit.
  2. Letter case is uppercase. Lowercase s is forbidden. The earlier PR #780 implementation that shipped [s] was a defect, fixed before merge.
  3. Letter font is JetBrains Mono 700. Inter is forbidden inside the mark. The bracket arms are SVG paths, not glyphs.
  4. Tile color follows the surface contract. Slate on light, light on dark, near-black on saturated slate. Never invent a fourth.
  5. The V0 in the wordmark is always demoted. Recoloring it to slate breaks the trademark hedge.
  6. New marks for new contexts require a DDR — they don't get added ad-hoc.

What this does not constrain:

  • The design system tokens (color, type, space) remain free to evolve in docs/design/design-system/.
  • Marketing surface treatments (slabs, hero blocks, partner deck templates) can iterate within the surface-contract rules.
  • The wordmark geometry — letter-spacing, weight, optical shift — is locked, but typography choices for product UI body copy are not.

Trademark caveat​

This decision was made on a designer-level similarity scan, not a formal trademark search. Counsel review is recommended before filing or external launch. See 03 · Trademark for the full hedge stance.

If counsel finds a conflict that requires a name change, this DDR is superseded.

References​