Skip to content

Latest commit

 

History

History
508 lines (391 loc) · 23.8 KB

File metadata and controls

508 lines (391 loc) · 23.8 KB

ReasonKit Brand Playbook

Classification: ULTIMATE MASTER SOURCE OF TRUTH Enforcement: HARD - ALL AI agents and projects MUST comply Last Updated: 2025-12-31 Supersedes: All other BRAND*.md files across RK-PROJECT


██████╗ ███████╗ █████╗ ███████╗ ██████╗ ███╗   ██╗██╗  ██╗██╗████████╗
██╔══██╗██╔════╝██╔══██╗██╔════╝██╔═══██╗████╗  ██║██║ ██╔╝██║╚══██╔══╝
██████╔╝█████╗  ███████║███████╗██║   ██║██╔██╗ ██║█████╔╝ ██║   ██║
██╔══██╗██╔══╝  ██╔══██║╚════██║██║   ██║██║╚██╗██║██╔═██╗ ██║   ██║
██║  ██║███████╗██║  ██║███████║╚██████╔╝██║ ╚████║██║  ██╗██║   ██║
╚═╝  ╚═╝╚══════╝╚═╝  ╚═╝╚══════╝ ╚═════╝ ╚═╝  ╚═══╝╚═╝  ╚═╝╚═╝   ╚═╝

              BRAND PLAYBOOK | DESIGNED, NOT DREAMED

Table of Contents

  1. Core Identity & Philosophy
  2. Voice & Tone Strategy
  3. Automated Style Enforcement
  4. Color System
  5. Typography
  6. Logo System: The Luminous Polyhedron
  7. Visual Language
  8. Asset Directory Structure
  9. Touchpoint-Specific Guidelines
  10. AI Agent Instructions

Core Identity & Philosophy

We are the architects of the post-prompting era.

While the industry "chats" with AI, hoping for the best, ReasonKit engineers it. We reject the chaotic unpredictability of raw LLMs in favor of auditable, structured, and deterministic protocols.

Our brand is not about "magic." It is about mastery. It evokes the feeling of a perfectly machined engine part, a clean compile, or a mathematical proof. It is the deep satisfaction of systems that work exactly as designed.

Attribute Value
Name ReasonKit
Tagline "Turn Prompts into Protocols"
Philosophy "Designed, Not Dreamed" / "High Logic > Lost Hopes"
Positioning "Structure Beats Intelligence"
Technical "Auditable Reasoning by Default"
Website https://reasonkit.sh
Mission Make AI reasoning structured, auditable, and reliable

Product Ecosystem

Project License Purpose Status
ReasonKit-core Apache 2.0 Structured reasoning engine OSS
ReasonKit-mem Apache 2.0 Memory/vector layer (optional) OSS
ReasonKit-web Apache 2.0 Web sensing MCP sidecar OSS
ReasonKit-pro Proprietary Advanced ThinkTools, enterprise Paid

Voice & Tone Strategy

The Vibe: The Lead Engineer who has seen it all. Calm, precise, slightly demanding, but ultimately reassuring because they have a plan.

The ReasonKit voice is unique. It is Competence Porn in text form. It doesn't sell you a dream; it hands you a blueprint.

The Voice Matrix

Dimension Description
Visionary but Grounded We talk about AGI and reasoning chains, but we ground every claim in Rust reliability and memory safety. We dream in code, not concepts.
Authoritative but Not Arrogant We don't say we're the best because we say so; we say it because we measured the variance reduction (85% → 28%). Show, don't just tell.
Technical but Accessible We assume intelligence but not specific context. We explain complex architectures with crystal clarity, respecting the reader's time.
Instructional vs. Conversational We don't "chat." We "direct." Our documentation reads like a flight manual, not a blog post.

The "Before & After" Examples

Context ❌ Generic AI Brand (Weak) ✅ ReasonKit Voice (Strong)
Value Prop "ReasonKit helps you get better answers from your AI models effortlessly." "ReasonKit forces LLMs to abandon probabilistic guessing in favor of structured, auditable reasoning chains."
Reliability "You can trust our system to be safe and secure." "Auditable by default. If it isn't in the logs, it didn't happen. We trade magic for metrics."
Speed "Blazing fast performance for your apps." "Core reasoning loops execute in <5ms. Rust supremacy means zero garbage collection pauses."
Error Handling "Oops! Something went wrong." "Reasoning chain diverged at Step 4. Circuit breaker activated. Retrying with temperature 0.1."

Key Lexicon & Banned Words

✅ USE THESE:

  • Protocol: Implies structure and rules.
  • Architecture: We build systems, not just prompts.
  • Determinism: The ultimate goal.
  • Variance Reduction: The measurable benefit.
  • Pipeline / Chain / Tree: Structural metaphors.
  • Inject / Extract / Synthesize: Active engineering verbs.

❌ BAN THESE:

  • Magic: There is no magic, only unexamined code.
  • Hope / Imagine: We don't hope; we design.
  • Seamless: Nothing is seamless. It is "integrated."
  • Game-changer: Show the delta, don't hype it.
  • Empower: Too vague. Use "enable" or "equip."

Automated Style Enforcement

To ensure this voice remains consistent across thousands of commits, we adhere to strict linting rules.

1. Vale Configuration (.vale.ini)

We use Vale to police our prose.

  • No Passive Voice: Active voice is responsible voice.
    • Bad: "The data was processed." (By whom? The data fairy?)
    • Good: "The ReasoningEngine processed the data."
  • No Weasel Words:
    • Error: "ReasonKit aims to improve..."
    • Fix: "ReasonKit improves..."
  • Sentence Length: Keep it punchy. If a sentence has more than 3 commas, break it.

2. Textlint Rules

  • textlint-rule-terminology: Enforces correct casing (ReasonKit, Rust, LLM, MCP).
  • textlint-rule-no-dead-link: A broken link is a broken promise.

3. ProofGuard in Documentation

Every claim in our documentation must be triangulated.

  • Rule: If you claim "20% faster," you must link to the benchmark.
  • Rule: If you claim "Secure by default," you must reference the specific security protocol (e.g., "Sandboxed WASM execution").

Color System

Source of Truth: reasonkit-site/main.css CSS variables Enforcement: NON-NEGOTIABLE - Use ONLY these colors

Primary Palette

Name Hex RGB CSS Variable Usage
Cyan (Primary) #06b6d4 rgb(6, 182, 212) --cyan CTAs, links, highlights, primary accent
Green (Success) #10b981 rgb(16, 185, 129) --green Success states, confirmations
Purple (Secondary) #a855f7 rgb(168, 85, 247) --purple Secondary accent, gradients
Pink (Tertiary) #ec4899 rgb(236, 72, 153) --pink Gradient endpoints, emphasis
Orange (Alert) #f97316 rgb(249, 115, 22) --orange Warnings, alerts
Yellow (Highlight) #fbbf24 rgb(251, 191, 36) --yellow Highlights, featured

Background & Text

Name Hex RGB CSS Variable Usage
Void Black #030508 rgb(3, 5, 8) --bg-void Primary background
Deep Black #0a0d14 rgb(10, 13, 20) --bg-deep Card backgrounds
Surface #111827 rgb(17, 24, 39) --bg-surface Elevated surfaces
Elevated #1f2937 rgb(31, 41, 55) --bg-elevated Hover states
White #f9fafb rgb(249, 250, 251) --text-primary Primary text
Text Secondary #9ca3af rgb(156, 163, 175) --text-secondary Secondary text
Text Muted #6b7280 rgb(107, 114, 128) --text-muted Muted text

Gradients

/* Hero Gradient (Cyan → Purple → Pink) */
--gradient-hero: linear-gradient(135deg, #06b6d4 0%, #a855f7 50%, #ec4899 100%);

/* Cyan-Purple Gradient */
--gradient-cyan-purple: linear-gradient(135deg, #06b6d4, #a855f7);

/* Glow Colors (30% opacity) */
--cyan-glow: rgba(6, 182, 212, 0.3);
--purple-glow: rgba(168, 85, 247, 0.3);
--pink-glow: rgba(236, 72, 153, 0.3);

Typography

Category Font Fallback Usage
Headlines Inter -apple-system, sans-serif H1, H2, hero text, body
Display Playfair Display Georgia, serif Special emphasis (optional)
Code JetBrains Mono SF Mono, Consolas, monospace Code, terminals, metrics

Font Weights

Weight Value Usage
Light 300 Large display text
Regular 400 Body text
Medium 500 Emphasis, subheadings
Semibold 600 Buttons, labels
Bold 700 Headlines, strong emphasis

Logo System: The Luminous Polyhedron

Canonical Logo: A glass-like hexagonal polyhedron representing multi-faceted, transparent reasoning.

Visual Metaphor

Element Meaning
Transparent facets Auditable, see-through reasoning
Multiple faces Multi-perspective analysis (GigaThink)
Glowing edges Structured protocols lighting the path
Central convergence All reasoning converges to clear conclusions

Logo Variants

Variant File Use Case
Icon Only logos/logo-icon.svg Favicons, app icons
Icon Transparent logos/logo-icon-transparent.svg Overlays, watermarks
Full Horizontal logos/logo-full-horizontal.svg Headers, footers
Full with Tagline logos/logo-with-tagline.svg Marketing materials
Stacked logos/logo-stacked.svg Square layouts, social
Wordmark logos/logo-wordmark.svg Text-only situations
Wordmark Dark logos/logo-wordmark-dark.svg Dark backgrounds
Wordmark Light logos/logo-wordmark-light.svg Light backgrounds

PNG Sizes Available

  • logo-icon-64.png, logo-icon-128.png, logo-icon-256.png, logo-icon-512.png
  • logo-full-400.png, logo-full-800.png
  • logo-wordmark-300.png, logo-wordmark-600.png

Usage Rules

  • Minimum clear space: 0.5x logo height
  • Minimum size: 32px height
  • ALWAYS use on dark backgrounds (Void Black preferred)
  • NEVER distort, rotate arbitrarily, or recolor
  • NEVER add drop shadows (logo has built-in glow effects)

Visual Language

Tree of Thoughts Motif

The core visual metaphor representing hierarchical reasoning:

  • Branching nodes = decision points
  • Glowing connections = reasoning pathways
  • Root = foundational axioms (BedRock)
  • Leaves = conclusions and outputs

Ambient Orbs

Floating, blurred orbs create depth:

  • Cyan, purple, pink colors
  • 15% opacity, 60px blur
  • 18-28s animation cycles

Glass Morphism

background: rgba(10, 15, 26, 0.8);
backdrop-filter: blur(12px);
border: 1px solid rgba(255, 255, 255, 0.05);

Asset Directory Structure

reasonkit-core/brand/                    # CANONICAL ASSET LOCATION
├── BRAND_PLAYBOOK.md                    # THIS FILE - Master source of truth
│
├── logos/                               # Logo variants (18 files)
│   ├── logo-icon.svg                    # Primary icon
│   ├── logo-icon-transparent.svg        # Transparent variant
│   ├── logo-full-horizontal.svg         # Full horizontal
│   ├── logo-stacked.svg                 # Stacked layout
│   ├── logo-with-tagline.svg            # With "Turn Prompts into Protocols"
│   ├── logo-wordmark*.svg               # Text-only variants
│   └── *.png                            # PNG exports (various sizes)
│
├── thinktools/                          # ThinkTool icons (12 files)
│   ├── icon-gigathink.svg               # GigaThink icon
│   ├── icon-laserlogic.svg              # LaserLogic icon
│   ├── icon-bedrock.svg                 # BedRock icon
│   ├── icon-proofguard.svg              # ProofGuard icon
│   ├── icon-brutalhonesty.svg           # BrutalHonesty icon
│   └── icon-*.svg                       # Other feature icons
│
├── badges/                              # Status badges (30 files)
│   ├── badge-rust.svg                   # Rust badge
│   ├── badge-apache-2.svg               # Apache 2.0 license
│   ├── badge-opensource.svg             # Open source badge
│   └── *.png                            # PNG badge exports
│
├── banners/                             # Hero/social banners (22 files)
│   ├── banner-hero.svg                  # Main hero banner
│   ├── banner-og-image.svg              # OG image template
│   ├── github-banner.svg                # GitHub social preview
│   ├── twitter-banner.svg               # Twitter/X banner
│   ├── linkedin-banner.svg              # LinkedIn banner
│   ├── og-social.png                    # Production OG image
│   ├── hero-tree.png                    # Hero tree visual
│   └── cognitive-engine.png             # Cognitive engine visual
│
├── patterns/                            # Background patterns (5 files)
│   ├── pattern-circuit.svg              # Circuit pattern
│   ├── pattern-hexagon.svg              # Hexagon pattern
│   ├── pattern-grid.svg                 # Grid pattern
│   ├── pattern-dots-hex.svg             # Dots pattern
│   └── circuit-void.svg                 # Void circuit pattern
│
├── diagrams/                            # Technical diagrams (13 files)
│   ├── reasontrace-topology.svg         # ReasonTrace visualization
│   ├── how-it-works.svg                 # How it works diagram
│   ├── deployment-modes.svg             # Deployment options
│   ├── thinktool-cards.svg              # ThinkTools overview
│   └── thinktools-chain.svg             # Chain visualization
│
├── favicons/                            # Favicon sizes (9 files)
│   ├── luminous-polyhedron.svg          # Master favicon SVG
│   ├── favicon.svg                      # Standard favicon
│   └── favicon-{16,32,48,64,128,256,512}.png
│
├── avatars/                             # Platform avatars (29 files)
│   ├── avatar-{192,400,800,1024}.png    # Generic sizes
│   ├── github-500.png                   # GitHub avatar
│   ├── twitter-400.png                  # Twitter/X avatar
│   ├── linkedin-400.png                 # LinkedIn avatar
│   └── {platform}-{size}.png            # Platform-specific
│
├── readme/                              # README images (14 files)
│   ├── reasonkit-core_hero.png          # Main README hero
│   ├── powercombo_process.png           # PowerCombo diagram
│   ├── chart_variance_reduction.png     # Performance chart
│   ├── terminal_mockup.png              # Terminal example
│   ├── architecture_diagram.png         # Architecture diagram
│   └── designed_not_dreamed.png         # Footer banner
│
└── launch-svg/                          # Launch SVG assets (5 files)
    ├── github-social-preview.svg        # GitHub preview
    ├── hero-reasoning-engine.svg        # Hero SVG
    ├── terminal-mockup.svg              # Terminal SVG
    ├── thinktools-overview.svg          # ThinkTools overview
    └── variance-reduction-chart.svg     # Chart SVG

Total Asset Count: ~160+ files


Touchpoint-Specific Guidelines

1. Website (ReasonKit.sh)

Primary touchpoint for users discovering ReasonKit.

Element Asset Location Notes
Favicon favicons/luminous-polyhedron.svg Luminous Polyhedron logo
Nav Logo Inline SVG (Luminous Polyhedron) 36x36px, with wordmark
Hero Banner banners/hero-tree.png or custom Cyan/purple gradient
OG Image banners/og-social.png 1200x630px
ThinkTool Icons thinktools/icon-*.svg Use in feature cards
Patterns patterns/pattern-*.svg Background accents

Website Asset Paths (RELATIVE):

/assets/brand/logo-*.svg
/assets/brand/favicon*.png
/og-image.png

2. README.md (ReasonKit-core)

Primary touchpoint for developers on GitHub.

Element Asset Path Notes
Hero Image ./brand/readme/reasonkit-core_hero.png Main banner
PowerCombo ./brand/readme/powercombo_process.png Process diagram
Variance Chart ./brand/readme/chart_variance_reduction.png Performance chart
Terminal ./brand/readme/terminal_mockup.png CLI example
Architecture ./brand/readme/architecture_diagram.png System diagram
Footer ./brand/readme/designed_not_dreamed.png Closing banner

README Image Reference Format:

![Description](./brand/readme/filename.png)

3. Other OSS READMEs (ReasonKit-mem, ReasonKit-web)

Copy relevant assets to each project's assets/ directory:

  • Use consistent hero image style
  • Reference reasonkit-core/brand/ for canonical assets
  • Maintain visual consistency

4. Social Media

Platform Size Asset
GitHub 1280x640 banners/github-banner.svg
Twitter/X 1500x500 banners/twitter-banner.svg
LinkedIn 1200x627 banners/linkedin-banner.svg
OG/Share 1200x630 banners/og-social.png
Avatar Per platform avatars/{platform}-{size}.png

AI Agent Instructions

For ALL AI Agents Working on ReasonKit

BRAND_COMPLIANCE_RULES:
  - ALWAYS use assets from: reasonkit-core/brand/
  - NEVER use off-brand colors (check hex values)
  - NEVER make unverified claims (ProofGuard discipline)
  - ALWAYS use Inter/JetBrains Mono fonts
  - ALWAYS use dark backgrounds (#030508 or #0a0d14)
  - NEVER generate AI text in images (add programmatically)

WHEN_CREATING_IMAGES:
  required_elements:
    - Cyan (#06b6d4) as primary color
    - Void black (#030508) background
    - Purple (#a855f7) and pink (#ec4899) accents
    - Tree/branching/neural network motifs
    - Clean, futuristic, premium aesthetic
    - NO embedded text (add via SVG/CSS later)

WHEN_WRITING_COPY:
  tone:
    - Authoritative (Competence Porn)
    - Technical (Engineered, not Dreamed)
    - Clear (No fluff)
    - Honest (Triangulated claims)
  avoid:
    - Buzzwords (Game-changer, Revolution)
    - Passive Voice (Own the action)
    - Unverified performance claims
    - Excessive exclamation marks
    - Vague or wishy-washy language (Aims to, Hopes to)

ASSET_SELECTION_PRIORITY:
  1. reasonkit-core/brand/           # ALWAYS use this first
  2. reasonkit-core/assets/img/      # For README-specific images
  3. reasonkit-site/assets/          # For website-specific assets
  4. DO NOT use rk-startup assets directly (they're copied to brand/)

Quick Asset Lookup

Need a logo?           → brand/logos/
Need a ThinkTool icon? → brand/thinktools/
Need a badge?          → brand/badges/
Need a banner?         → brand/banners/
Need a pattern?        → brand/patterns/
Need a diagram?        → brand/diagrams/
Need a favicon?        → brand/favicons/
Need an avatar?        → brand/avatars/
Need README images?    → brand/readme/
Need launch SVGs?      → brand/launch-svg/

Version History

Version Date Changes
2.0.0 2026-01-08 MAJOR: Complete "Brand Voice" overhaul. Added "Competence Porn" strategy, Voice Matrix, and Automated Enforcement sections. Updated philosophy to "Designed, Not Dreamed."
1.0.0 2025-12-31 Initial BRAND_PLAYBOOK.md - Consolidated all brand assets.

Enforcement

This document is the SINGLE SOURCE OF TRUTH for all ReasonKit brand identity.

All AI agents, projects, documentation, and marketing materials MUST:

  1. Use assets from reasonkit-core/brand/ directory
  2. Follow color, typography, and visual guidelines exactly
  3. Apply ProofGuard discipline to all claims
  4. Maintain consistency across all touchpoints

Deviations require explicit documentation and approval.


"Designed, Not Dreamed" | https://reasonkit.sh