ornament-style-modern
정보
이 스킬은 사이버펑크나 베이퍼웨이브 같은 현대적이고 공상적인 미학을 활용하여 장식적 패턴을 생성합니다. 색맹 친화적이고 지각적으로 균일한 버미디스(Viridis)와 인페르노(Inferno) 같은 색상 척도를 사용하여 접근성을 우선시합니다. 현대적인 장르 특화 디자인을 제작하거나 색맹 접근성이 요구되는 경우에 사용하세요.
빠른 설치
Claude Code
추천npx skills add pjt222/agent-almanac -a claude-code/plugin add https://github.com/pjt222/agent-almanacgit clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/ornament-style-modernClaude Code에서 이 명령을 복사하여 붙여넣어 스킬을 설치하세요
문서
Ornament Style — Modern
Design ornamental patterns using modern and speculative aesthetics with colorblind-accessible color scales. This skill breaks free from historical period constraints — anachronism is encouraged, hybrid motifs are welcome, and the palette system is built on perceptually uniform color scales designed for universal accessibility.
This is the "unleashed" companion to ornament-style-mono and ornament-style-color. Where those skills ground every decision in art historical fidelity, this skill grounds decisions in genre coherence, color accessibility, and artistic freedom.
When to Use
- Creating ornamental designs in modern, speculative, or genre-specific aesthetics (cyberpunk, solarpunk, brutalist, etc.)
- Designing patterns that must be colorblind-accessible using scientifically validated color scales
- Exploring hybrid motifs that combine historical ornament with contemporary visual language
- Producing ornament for digital contexts (UI decoration, game assets, screen-based media) where historical authenticity is not the goal
- Generating decorative imagery where CVD (Color Vision Deficiency) safety is a requirement
- Creating purely abstract or algorithmic ornament with no historical reference
- Combining motif traditions across cultures and periods without concern for anachronism
Inputs
- Required: Genre / aesthetic direction (or "surprise me" for random selection, or "no genre" for pure abstract)
- Required: Application context (border, medallion, frieze, panel, tile, standalone motif, UI element)
- Optional: Color scale preference (viridis, cividis, inferno, magma, plasma, etc.) or custom palette
- Optional: CVD type to optimize for (protanopia, deuteranopia, tritanopia, or "all")
- Optional: Specific motif elements (circuit traces, organic growth, geometric lattice, etc.)
- Optional: Rendering style (digital art, holographic, neon sign, 3D render, glitch art, etc.)
- Optional: Historical hybrid elements (e.g., "Gothic tracery + circuit board")
- Optional: Target resolution and aspect ratio
- Optional: Seed value for reproducible generation
Procedure
Step 1: Select Genre / Aesthetic
Choose a modern or speculative aesthetic as the visual foundation. Unlike historical periods, genres are fluid — mixing and hybridization are encouraged.
Modern and Speculative Aesthetics:
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
| Genre | Visual Character | Motif Language | Color Tendency |
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
| Cyberpunk | Neon-on-dark, circuit, | Circuit traces, kanji, hexagons, | Neon cyan/magenta on black |
| | tech, glitch | fractured glass, data streams | |
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
| Hard Sci-Fi | Clean, technical, | Engineering diagrams, orbital | Cool metallics, deep space |
| | precise | paths, crystalline lattice | blue |
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
| Solarpunk | Organic-tech fusion, | Leaf/vine + solar panel, living | Greens, warm amber, |
| | verdant | architecture, moss on circuit | sunlight |
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
| Biopunk | Organic, visceral, | DNA helix, cell structure, | Deep organics, |
| | grown | mycelium, coral, nerve networks | bioluminescent |
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
| Brutalist | Raw, monumental, | Concrete texture, massive | Grays, concrete, |
| | geometric | geometry, exposed grid, slab | industrial |
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
| Art Deco Revival | Geometric elegance, | Sunburst, chevron, ziggurat, fan, | Gold, black, cream, |
| | luxury | stepped forms | emerald |
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
| Vaporwave | Retro-digital, surreal | Classical busts (glitched), grid, | Pink/teal/purple gradients |
| | | gradient, marble | |
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
| Retrofuturism | 1950s-60s future vision | Atomic/Space Age, streamline, ray | Chrome, turquoise, coral |
| | | gun, fin, orbit | |
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
| Generative / | Mathematical, | Voronoi, reaction-diffusion, | Any scale -- often viridis |
| Algorithmic | procedural | L-system, noise field, fractal | family |
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
| Afrofuturism | Ancestral + futuristic | Adinkra, kente geometry, cosmic, | Rich earth + metallic + |
| | | mask, constellation | electric |
+-----------------------+--------------------------+-----------------------------------+-----------------------------+
Additional supported modes:
- Freeform user-defined aesthetic: describe any visual direction not in the table
- Historical hybrid: combine a historical period with a modern genre (e.g., "Byzantine cyberpunk," "Islamic generative")
- No genre / pure abstract: ornament driven entirely by structure, color scale, and composition with no genre narrative
- If the user specified a genre, confirm and note its visual character, motif language, and color tendency
- If "surprise me," select randomly — weight toward genres with strong ornamental potential (cyberpunk, solarpunk, generative, Art Deco Revival)
- If a historical hybrid is requested, identify the historical motif vocabulary and the modern rendering/mood overlay
- Note the application context (digital screen, print, physical object) as this affects rendering choices
Got: A clearly identified genre (or hybrid) with its characteristic visual language understood. For hybrids, both source traditions should be articulated.
If fail: If the user requests a genre not in the table, research its visual conventions using WebSearch for "[genre] aesthetic visual design motifs" and construct an equivalent entry. The key elements to identify are: visual character, typical motifs, and color tendency.
Step 2: Select Color Strategy
Choose between a colorblind-accessible scale or a custom palette. Colorblind scales are the recommended default.
Path A: Colorblind-Accessible Scale (Recommended)
Select from perceptually uniform color scales designed for universal readability:
Colorblind-Accessible Color Scales:
+------------+---------------+--------------------+-----------------------------+-------------------------+
| Scale | Type | CVD Safe | Character | Best For |
+------------+---------------+--------------------+-----------------------------+-------------------------+
| Viridis | Sequential | All 3 types | Blue-green-yellow, balanced | Default recommendation |
+------------+---------------+--------------------+-----------------------------+-------------------------+
| Cividis | Sequential | Deutan + Protan | Blue-to-yellow, muted | Maximum CVD safety |
| | | optimized | | |
+------------+---------------+--------------------+-----------------------------+-------------------------+
| Inferno | Sequential | All 3 types | Black-red-yellow-white, hot | Dramatic, high contrast |
+------------+---------------+--------------------+-----------------------------+-------------------------+
| Magma | Sequential | All 3 types | Black-purple-orange-yellow | Moody, volcanic |
+------------+---------------+--------------------+-----------------------------+-------------------------+
| Plasma | Sequential | All 3 types | Purple-pink-orange-yellow | Vivid, energetic |
+------------+---------------+--------------------+-----------------------------+-------------------------+
| Turbo | Rainbow-like | Moderate | Full spectrum, perceptual | Many distinct colors |
+------------+---------------+--------------------+-----------------------------+-------------------------+
| Mako | Sequential | All 3 types | Dark blue to light teal | Cool, oceanic |
+------------+---------------+--------------------+-----------------------------+-------------------------+
| Rocket | Sequential | All 3 types | Dark to light via warm | Warm, ember-like |
| | | | tones | |
+------------+---------------+--------------------+-----------------------------+-------------------------+
| Okabe-Ito | Categorical | All 3 types | 8 distinct colors | Discrete elements, max |
| | | | | distinction |
+------------+---------------+--------------------+-----------------------------+-------------------------+
CVD Types and Impact:
+-----------------------------+----------+-----------------+------------------------------+
| CVD Type | Affects | Confuses | Safe Scales |
+-----------------------------+----------+-----------------+------------------------------+
| Protanopia (red-blind) | ~1% male | Red/green | viridis, cividis, inferno |
+-----------------------------+----------+-----------------+------------------------------+
| Deuteranopia (green-blind) | ~5% male | Red/green | viridis, cividis, inferno |
+-----------------------------+----------+-----------------+------------------------------+
| Tritanopia (blue-yellow) | ~0.01% | Blue/yellow | inferno, magma (with care) |
+-----------------------------+----------+-----------------+------------------------------+
Translating a continuous scale to an ornamental palette: Sample 3-5 colors at evenly spaced intervals along the scale. For a 5-color palette from viridis:
- Position 0.0: deep purple (#440154)
- Position 0.25: blue-violet (#31688e)
- Position 0.5: teal green (#35b779)
- Position 0.75: yellow-green (#90d743)
- Position 1.0: bright yellow (#fde725)
Assign roles using the 60/30/10 framework from ornament-style-color: dominant (largest area), secondary (supporting), accent (focal points).
Path B: Custom Palette
Freeform palette selection with optional CVD simulation check:
- Define 3-5 colors with named roles
- Optionally validate against CVD types using WebSearch for "CVD color simulator [colors]"
- Note any CVD risk and mitigations (e.g., using texture or pattern in addition to color)
Got: A palette of 3-5 named colors with roles, either sampled from a named scale or custom-defined, with CVD compatibility noted.
If fail: If unsure, use viridis with 3-color sampling (deep purple, teal, yellow). This is the most universally accessible and visually balanced default.
Step 3: Analyze Motif Structure
Understand the structural grammar of the chosen motif, using the same framework as the historical skills but with explicit permission for modern compositional techniques.
-
Perform structural analysis:
- Symmetry type: bilateral, radial, translational, point, or glitch-broken, procedural asymmetry, pseudo-random
- Geometric scaffold: circle, rectangle, triangle, band, or voronoi cell, fractal, reaction-diffusion field
- Fill pattern: solid, line-filled, open, mixed, or gradient, noise texture, data-driven
- Edge treatment: clean, organic, interlocking, or glitched, dissolving, pixel-stepped
-
Add color-to-structure mapping:
- Which structural elements receive which colors from the selected scale?
- Does color follow form (each shape gets one color) or flow (gradient across structural boundaries)?
- Where does the scale's brightest/lightest color appear? (Typically focal points)
- What color is the ground/background?
-
Modern composition techniques (unique to this skill):
- Hybrid motifs: historical ornament structure + modern rendering (e.g., Gothic tracery rendered as circuit board)
- Non-traditional symmetry: deliberate symmetry breaks, glitch artifacts, procedural variation within a repeating pattern
- Layered compositions: ornament over texture, transparency effects, depth-of-field blur
- Meta-ornament: ornamental patterns composed of smaller ornamental patterns (fractal nesting)
Got: A structural description with explicit color assignments and any modern composition techniques identified.
If fail: If the motif structure is unclear for a modern genre, anchor to the genre's real-world visual precedents. Cyberpunk circuit traces follow band scaffold with translational symmetry. Generative/algorithmic uses radial or field-based scaffold. The motif language may be novel but the structural grammar is universal.
Step 4: Construct Modern Prompt
Build the text prompt for Z-Image generation, using the modern prompt template.
Prompt Template:
[Rendering style] of [genre]-inspired ornamental design,
[motif description], [color scale or palette],
[composition type], [mood/atmosphere],
[application context], [additional qualifiers]
Modern Rendering Styles:
digital art— clean digital rendering, screen-ready, smooth gradientsholographic— iridescent, light-diffracting, multi-angle color shiftneon sign— glowing lines on dark ground, light bloom, hot edges3D render— volumetric, lit, material quality, depthglitch art— digitally corrupted, scan-line artifacts, color channel splitvector graphic— flat, clean edges, scalable feel, geometric precisionscreen print— limited color, registration marks, ink quality, tactilelaser etched— precise, metallic surface, ablation marks, technicalgenerative art— procedural, algorithmic, mathematical precisionconcept art— painterly, atmospheric, narrative, cinematic lighting
Color Scale in Prompts: Translate the scale name into descriptive color language that the model can interpret:
- Viridis: "deep purple transitioning through teal green to bright yellow"
- Cividis: "steel blue transitioning to golden yellow"
- Inferno: "black through deep red and orange to bright yellow-white"
- Magma: "black through dark purple and burnt orange to pale yellow"
- Plasma: "deep indigo through magenta and orange to bright yellow"
- Mako: "deep navy blue transitioning to light aqua teal"
- Rocket: "dark brown-black through brick red to pale cream"
Example Prompts:
neon sign ornamental design inspired by cyberpunk aesthetic, circuit trace patterns and hexagonal grid, deep purple and teal green and bright yellow colors (viridis palette), repeating border frieze, electric and atmospheric, dark background with glowing elementsdigital art of solarpunk-inspired ornamental panel, vine and leaf motifs intertwined with solar cell geometry, steel blue to golden yellow gradient (cividis palette), vertical panel composition, warm and hopeful atmosphere, organic-technology fusiongenerative art ornamental tile, algorithmic reaction-diffusion pattern, dark purple through burnt orange to pale yellow (magma palette), square repeating unit, mathematical and volcanic, procedural organic quality3D render of Art Deco Revival ornamental medallion with brutalist influence, sunburst and ziggurat geometry in raw concrete and gold, deep indigo through magenta to bright yellow (plasma palette), radial symmetry, monumental elegance
Got: A prompt of 25-50 words that specifies rendering style, genre, motif, color scale/palette, composition, and mood.
If fail: If the prompt produces colors that do not match the intended scale, front-load the color description. Instead of mentioning the scale name, describe the actual colors at the start: "deep purple, teal green, and bright yellow ornamental design..." Z-Image weights earlier prompt tokens more heavily.
Step 5: Configure Generation Parameters
Select resolution and generation parameters.
Resolution by Application:
+--------------------+---------------------+----------------------------------------+
| Application | Recommended | Rationale |
+--------------------+---------------------+----------------------------------------+
| Medallion / Roundel| 1024x1024 (1:1) | Radial symmetry needs square |
+--------------------+---------------------+----------------------------------------+
| Tile / Repeat Unit | 1024x1024 (1:1) | Square for seamless tiling |
+--------------------+---------------------+----------------------------------------+
| Horizontal Frieze | 1280x720 (16:9) | Wide format for running border |
+--------------------+---------------------+----------------------------------------+
| Vertical Panel | 720x1280 (9:16) | Portrait format for columns |
+--------------------+---------------------+----------------------------------------+
| Wide Border | 1344x576 (21:9) | Ultrawide for architectural |
+--------------------+---------------------+----------------------------------------+
| UI Element | 1152x896 (9:7) | Balanced landscape for screen use |
+--------------------+---------------------+----------------------------------------+
| Large Detail | 1536x1536 (1:1) | Higher res for fine gradient work |
+--------------------+---------------------+----------------------------------------+
- Select resolution based on application context
- Set
stepsto 10-12 (color gradient detail and scale fidelity benefit from more steps) - Set
shiftto 3 (default) or 4 for neon-on-dark styles that benefit from higher contrast - Choose
random_seed: truefor exploration orrandom_seed: falsewith a specific seed for reproducibility - Record all parameters for documentation
Got: A complete parameter set. Gradient-based scales need 10+ steps for smooth color transitions.
If fail: If unsure, use 1024x1024 at 10 steps with shift 3. Increase to shift 4 only for neon/glowing/high-contrast styles.
Step 6: Generate Image
Invoke the Z-Image MCP tool to produce the ornament.
- Call
mcp__hf-mcp-server__gr1_z_image_turbo_generatewith:prompt: the constructed prompt from Step 4resolution: from Step 5steps: from Step 5 (recommend 10-12)shift: from Step 5random_seed: from Step 5seed: specific seed ifrandom_seedis false
- Record the returned seed value for reproducibility
- Note the generation time
Got: A generated image with recognizable ornamental structure and visible color gradient/palette. The color may not perfectly match the specified scale — this is addressed in evaluation.
If fail: If the MCP tool is unavailable, verify that hf-mcp-server is configured (see configure-mcp-server or troubleshoot-mcp-connection). If the generated image is entirely abstract with no ornamental structure, the prompt needs more specific structural language — return to Step 4. If colors are completely wrong, front-load the color names in the prompt.
Step 7: Evaluate Design
Assess the generated image against five criteria adapted for modern ornament.
Modern Ornament Evaluation Rubric:
+---------------------+------------------------+-------------------------------------------+
| Criterion | Replaces (from color) | Evaluation Questions |
+---------------------+------------------------+-------------------------------------------+
| 1. Genre Coherence | Period Accuracy | Does it feel like the specified genre? |
| | | Would someone familiar with the genre |
| | | recognize the aesthetic? |
+---------------------+------------------------+-------------------------------------------+
| 2. Color Scale | Palette Match | Does the color gradient/palette |
| Fidelity | | approximate the chosen scale? Are the |
| | | key colors from the scale present? |
+---------------------+------------------------+-------------------------------------------+
| 3. Accessibility | (new criterion) | Would this be distinguishable under the |
| | | target CVD type? Do elements rely solely |
| | | on color or also on shape/texture? |
+---------------------+------------------------+-------------------------------------------+
| 4. Composition | Form-Color Balance | Does the ornamental structure read |
| Quality | | clearly? Does color clarify or obscure |
| | | the motif? |
+---------------------+------------------------+-------------------------------------------+
| 5. Rendering Style | Rendering Style | Does it match the specified rendering |
| | | technique? Does a "neon sign" glow? |
| | | Does "glitch art" show artifacts? |
+---------------------+------------------------+-------------------------------------------+
- Score each criterion: Strong (clearly meets), Adequate (partially meets), Weak (does not meet)
- Note specific observations for each criterion
- If 4+ criteria score Strong, the design is successful
- If 2+ criteria score Weak, return to Step 4 for prompt refinement
Got: A scored evaluation with specific observations. Modern styles with gradient color scales are harder to control than flat historical palettes — expect Adequate on first generation for color scale fidelity.
If fail: If most criteria score Weak, the prompt may need fundamental restructuring. Common fixes: move color descriptions to the very beginning of the prompt, simplify to fewer colors (3 instead of 5), strengthen the genre-specific language, increase steps to 12.
Step 8: Iterate or Finalize
Refine the design through targeted iteration or accept the result.
Modern-Specific Iteration Strategies:
- Scale sampling shift: If the viridis-derived palette is too compressed, sample from different positions on the scale (e.g., skip the middle, use endpoints + one off-center point)
- Genre amplification: If the genre is not coming through, add genre-specific keywords: "cyberpunk neon circuitry" instead of just "cyberpunk"
- Color front-loading: Put specific color descriptions at the very start of the prompt
- Seed-locked color tuning: Keep the seed, change only the color description to adjust palette while maintaining composition
- Rendering correction: Strengthen the rendering style with material-specific language: "glowing neon tubes on matte black surface" instead of just "neon sign"
- Accessibility enhancement: If CVD evaluation is weak, increase contrast between adjacent elements and add structural differentiation (texture, pattern, size) alongside color
Iteration Budget: Limit to 3 iterations per design concept.
- If the evaluation in Step 7 indicates specific weaknesses, apply the corresponding correction strategy
- Regenerate using Step 6
- Re-evaluate using Step 7
- Accept when 4+ criteria score Strong or iteration budget is exhausted
Got: Improved genre coherence and color fidelity after 1-2 iterations. Perfect scale reproduction is unlikely — aim for "recognizably in the right color family with correct progression direction."
If fail: If iteration is not converging, the color scale may be too subtle for the model to reproduce as a gradient. Simplify by sampling fewer colors from the scale (3 instead of 5) and naming them explicitly. Alternatively, accept the closest approximation and note the deviation in documentation.
Step 9: Document the Design
Create a complete record of the final design for reproducibility and reference.
- Record the following:
- Genre: Genre/aesthetic name and any hybrid elements
- Motif: Primary motif(s) used and their structural grammar
- Rendering Style: Digital art, neon sign, glitch art, 3D render, etc.
- Color Scale: Scale name and sample points, or custom palette
- If scale: [scale name], sampled at positions [0.0, 0.25, 0.5, 0.75, 1.0]
- If custom: each color with role and approximate hex value
- CVD Compatibility: Target CVD type(s) and assessment
- Color Roles (60/30/10):
- Dominant: [color name from scale] (~hex) — 60%
- Secondary: [color name from scale] (~hex) — 30%
- Accent: [color name from scale] (~hex) — 10%
- Final Prompt: The exact prompt that produced the accepted image
- Seed: The seed value for reproduction
- Resolution: The resolution used
- Steps/Shift: Generation parameters
- Evaluation: Brief notes on the five rubric criteria scores
- Iterations: Number of iterations and key changes made
- Note genre coherence observations (what worked, what the model interpreted differently)
- Note CVD-specific observations (elements that rely on color alone vs. color + structure)
- Suggest potential applications and adaptation notes
Got: A reproducible record with full color scale documentation and CVD compatibility assessment.
If fail: If full documentation feels excessive, at minimum record the final prompt, seed, color scale name, and CVD compatibility status. These allow reproduction and accessibility verification.
Validation
- A genre or aesthetic direction was selected (or explicit "no genre" for pure abstract)
- Color strategy was chosen: named colorblind scale or custom palette with CVD check
- If using a colorblind scale, sample points were identified and roles assigned
- CVD compatibility was assessed for the target audience
- Motif structure was analyzed with color-to-structure mapping
- Prompt includes explicit color descriptions (not just scale name) and genre-specific language
- Prompt specifies a modern rendering style appropriate to the genre
- Resolution matches the application context
- Steps set to 10+ for gradient/color fidelity
- Generated image was evaluated against the 5-point modern rubric (including accessibility criterion)
- Seed value was recorded for reproducibility
- Final design is documented with prompt, seed, scale/palette, CVD notes, and parameters
Pitfalls
- Using scale names in prompts: Z-Image does not know "viridis" — translate to descriptive colors: "deep purple through teal green to bright yellow." The scale name is for your documentation, the color words are for the prompt
- Ignoring CVD beyond scale selection: Choosing a CVD-safe scale is necessary but not sufficient. If the ornament relies on distinguishing adjacent colors in the scale without structural differentiation (shape, texture, size), it may still be inaccessible. Use redundant visual coding
- Genre without structure: "Cyberpunk ornament" is too vague. Specify the motifs: "cyberpunk circuit trace border with hexagonal nodes." Genre is atmosphere; motifs are structure. You need both
- Too many scale samples: Sampling 7+ points from a continuous scale creates a muddy gradient in generation. 3-5 sample points produce cleaner results with better scale fidelity
- Neglecting the dark ground: Many modern genres (cyberpunk, neon, vaporwave) assume a dark background. Failing to specify "on dark background" or "on black ground" produces washed-out results with bright scales
- Insufficient steps for gradients: Gradient-based color scales need more inference steps than flat historical palettes. Using 8 steps for scale-based color work produces banded or imprecise transitions. Use 10-12
- Forcing historical fidelity in a modern skill: This skill is not
ornament-style-color. If you find yourself policing anachronism or insisting on period-authentic pigments, switch to the historical skills. Here, a Byzantine motif rendered as a cyberpunk neon sign is not an error — it is the point
Related Skills
ornament-style-mono— the monochrome foundation skill; useful for establishing motif structure before adding modern color treatmentornament-style-color— the historical color companion; use when period-authentic palettes and art historical fidelity are required instead of modern aestheticsreview-web-design— color theory and accessibility principles apply to ornamental color compositionreview-ux-ui— WCAG color contrast guidelines are relevant when ornament is used in UI contextsmeditate— focused attention and visualization practices can inform abstract pattern development
GitHub 저장소
연관 스킬
content-collections
메타이 스킬은 콘텐츠 콜렉션(Content Collections)을 위한 프로덕션 검증된 설정을 제공합니다. 콘텐츠 콜렉션은 Markdown/MDX 파일을 Zod 검증이 포함된 타입 안전한 데이터 콜렉션으로 변환해주는 TypeScript 최우선 도구입니다. 블로그, 문서 사이트 또는 콘텐츠 중심의 Vite + React 애플리케이션을 구축할 때 타입 안전성과 자동 콘텐츠 검증을 보장하기 위해 사용하세요. Vite 플러그인 구성과 MDX 컴파일부터 배포 최적화 및 스키마 검증에 이르기까지 모든 것을 다룹니다.
polymarket
메타이 스킬은 개발자들이 Polymarket 예측 시장 플랫폼을 활용한 애플리케이션을 구축할 수 있도록 지원하며, 거래 및 시장 데이터를 위한 API 통합 기능을 포함합니다. 또한 WebSocket을 통한 실시간 데이터 스트리밍을 제공하여 실시간 거래와 시장 활동을 모니터링할 수 있습니다. 이를 통해 거래 전략을 구현하거나 실시간 시장 업데이트를 처리하는 도구를 생성하는 데 활용할 수 있습니다.
creating-opencode-plugins
메타이 스킬은 개발자들이 명령어, 파일, LSP 작업 등 25개 이상의 이벤트 유형에 연결되는 OpenCode 플러그인을 만들 수 있도록 돕습니다. JavaScript/TypeScript 모듈을 위한 플러그인 구조, 이벤트 API 명세, 구현 패턴을 제공합니다. OpenCode AI 어시스턴트의 라이프사이클을 사용자 정의 이벤트 기반 로직으로 가로채거나, 모니터링하거나, 확장해야 할 때 사용하세요.
sglang
메타SGLang은 RadixAttention 프리픽스 캐싱을 활용하여 JSON, 정규식, 에이전트 워크플로우를 위한 고속 구조화 생성에 특화된 고성능 LLM 서빙 프레임워크입니다. 특히 반복되는 프리픽스가 있는 작업에서 상당히 빠른 추론 속도를 제공하여 복잡한 구조화 출력 및 다중 턴 대화에 이상적입니다. 제약 디코딩이 필요하거나 광범위한 프리픽스 공유가 있는 애플리케이션을 구축할 때는 vLLM과 같은 대안보다 SGLang을 선택하십시오.
