SPB Git

Repository Public License

brand-guidelines.md 7.9 KB

# Brand guidelines

# The ORIX family

ORIX apps share one design package (Packages/ORIXDesign): the same neutral palette, spacing scale, radii, type styles, motion and components. What identifies the family visually is the ORIX ring — four rounded arcs, one per quadrant, with a gap on each axis — and a flat, calm interface with a single level of elevation.

Each app brings exactly one thing of its own: an accent pair (ORIXAccent, a primary and a secondary colour). Everything else is inherited. An ORIX app should be recognisable as family at a glance and distinguishable by its accent and by what sits inside the ring.

# ORIX Secure

Name ORIX Secure
Tagline Know Your Device. Protect What Matters.
Accent, primary violet — #7C3AED (light), #A78BFA (dark)
Accent, secondary mint, the "guard ring" — #0E9F6E (light), #34D399 (dark)
Mark the ORIX ring, an unbroken mint guard ring inside it, and a solid core at the centre

The mark reads from the outside in: the family ring; a closed ring that guards; the thing being protected. The guard ring is the only unbroken circle in the mark and the only element in mint. It is described in icon-design.md.

Violet carries identity and interaction (buttons, glyphs, tint). Mint is used sparingly: the guard ring, and the end of the progress gradient. Neither accent colour ever means "good" or "bad": state colours are separate tokens (see design-system.md).

# Name usage

  • Write ORIX Secure: "ORIX" in capitals, a space, "Secure" with a capital S. This is the display name in Info.plist and it is never translated (Scripts/localize.py marks it as not translatable).
  • Not "Orix Secure", "ORIXSecure", "ORIX SECURE" or "Orix".
  • ORIXSecure without a space is a technical identifier only (target, scheme, bundle name, repository folder). It does not appear in text a user reads.
  • Do not shorten the product to "ORIX" in running text: ORIX is the family.
  • Do not attach category words that claim more than the app does: not "ORIX Secure Antivirus", "ORIX Secure Cleaner", "ORIX Secure Protection". The app's own capability statement says "ORIX Secure is not an antivirus".
  • The tagline is written exactly as above, in title case with two full stops, and is used as a pair. It describes what the user does with the app (know, protect); it does not promise that the app protects the device.

# Tone of voice for security text

Security apps commonly earn attention by frightening people. ORIX Secure does the opposite, and the rules below are enforced by tests, not left to taste.

# Rules

  1. Calm. State what was found in the register of a measurement, not an alarm. No exclamation marks, no capitals for emphasis, no countdowns, no "act now".
  2. Evidence first. Every finding carries its evidence (label, value, source), how it was established (read from a documented API, parsed from the data, local heuristic, confirmed by the user, not verifiable) and what was not inspected. A statement without evidence is not shipped; the engine tests fail on a finding with empty evidence.
  3. Say whose knowledge it is. "Verified" is reserved for what the app read itself. What the user answered is "Confirmed by you". What cannot be read is "Not available", with the reason.
  4. No verdicts. The app describes signals; it does not pronounce a link, file, network or device good or bad. "No signals found" is not "safe", and the text says so ("not proof that a link is harmless").
  5. Heuristics are named. A result from a rule of thumb is labelled "Heuristic" and never presented with the weight of a measurement.
  6. State the limits where the result is shown, not in a legal page: what no app can see is listed next to the result (LimitsCard).
  7. A count, never a score. The overview shows how many items were checked and with what result. It does not compute a security score or a percentage, and says so.
  8. Recommend a concrete action. An attention finding says what to do and where in Settings, in one or two sentences.
  9. Red is rare. Only a high-severity attention finding is red. Moderate and low are amber; unknown is grey; a user confirmation is blue.

# Forbidden words

These lists are copied from the test targets. A finding, summary, methodology text or type name containing one of them fails the engine tests.

Scope Never written
Device, checklist, privacy, investigations "100% secure", "malware-free", "no malware detected", "virus", "infected", "hacked", "your device is safe", "is safe to"
Links and QR codes "safe", "unsafe", "safely", "malicious", "clean", "infected", "virus", "viruses"
Files and archives "safe", "malicious", "infected", "virus", "malware detected", "clean"
Network "hacked", "safe", "secure network"

The UI tests add screen-level checks: no "100%", no "No malware", no "is safe", no invented "°C", no button containing "boost".

By extension, in any text written for the product (store listing, release notes, support replies), avoid: "scan for viruses", "threat detected", "protected", "secured", "clean your phone", "speed up", "boost", "optimise memory", and any claim that the app detects malware or spyware. The word "malware" appears only in statements of what the app does not do ("No app can scan an iPhone for malware").

# Examples

Instead of Write
"Your device is safe" "12 of 16 items checked. This shows what was checked, not how secure the device is."
"Dangerous link!" "Elevated — the registrable domain is example.net; 'apple.com' appears only as a sub-domain."
"This link is safe" "No signals found. This is not proof that a link is harmless."
"Infected file" "The name ends in .pdf but the content is a Windows program."
"Two-factor authentication: verified" "Two-factor authentication: confirmed by you"
"Battery temperature: 31 °C" (not readable) "Processor temperature: not available — iOS does not expose it to apps."
"You may have been hacked" "Supported by your answers: the password was entered on a page reached from the link. Change that password."
"Clean 2.3 GB now" "2 extra copies in 1 group. Deleting moves them to Recently Deleted."

The right-hand column illustrates the register; the exact strings in the app live in the code and the string catalog.

# Do and don't

Do

  • Use the mark on the violet gradient tile, or as a single-colour template mark on a plain background.
  • Keep clear space around the mark of at least the width of one arc stroke.
  • Use violet for interaction and identity, state colours only for states.
  • Pair every colour-coded state with its symbol and its words.
  • Describe features by what they inspect: "analyses a link's structure", "reads the archive's directory without extracting it", "compares photo fingerprints on the device".
  • Say "on this device" when it is true: analysis is local and nothing is uploaded.
  • Show what was not inspected.

Don't

  • Don't use shields with check marks, padlocks, radar sweeps, bugs, skulls or warning triangles as brand imagery. The mark is the ring; exclamationmark.triangle is a state symbol, not decoration.
  • Don't use red for branding, backgrounds, badges or marketing. Red belongs to high-severity findings.
  • Don't recolour the guard ring, break it, or remove it: an ORIX ring without the guard ring is another app's mark.
  • Don't rotate the mark, add effects to it, or place the full-colour icon on a busy photograph.
  • Don't show a score, a percentage of "security", a gauge or a "protected" badge.
  • Don't animate to create urgency (pulsing alerts, shaking, flashing).
  • Don't promise outcomes the app cannot verify: freed space before deletion is confirmed, "all threats removed", "your network is secure".
  • Don't translate or decline the name.
Files main
Menu

Theme