Tag: prompt-engineering
Brand-Consistent Illustration Set Generation Skill
This skill manages the challenge of maintaining brand style across multiple assets by injecting a structured brand bundle, including palette constraints and style references. It supports progressive style locking, allowing initial human-gat…
Asset Generation Diagnosis and Repair Skill
This skill diagnoses and repairs failures in automated asset generation by mapping validation error codes to specific repair primitives. It implements a retry budget and routing logic to resolve issues such as palette drift, transparency er…
UI Brief to Designer Mockup Prompt Generator
This skill translates a functional UI brief into a highly constrained, paste-ready prompt for advanced image generation models. It intelligently selects the optimal model and enforces anti-slop design rules to ensure the output provides gen…
Model-Specific Image Prompt Dialect Generator
This skill translates a normalized asset brief into a highly specific prompt dialect tailored for various text-to-image models. It intelligently handles model quirks, such as token budgets, negative prompt syntax, text-in-image ceilings, an…
CLI tool for deep LLM code review
This CLI tool facilitates deep, context-aware code analysis by bundling a prompt with selected project files for a single LLM pass. It supports multiple engines (browser/API) and manages long-running sessions for complex debugging or refact…
LLM Prompt Engineering and Optimization Skill
This skill focuses on designing, testing, and optimising LLM prompts and system instructions. It covers techniques including chain-of-thought, few-shot prompting, and token budget optimisation for various models.
Structured Image Prompt Engineering Skill
Implements a structured five-section framework to transform vague user requests into precise, text-accurate prompts for models like GPT Image 2 and Flux. It uses specific rules for typography, layout preservation, and anti-slop to ensure hi…
Agent Skill Authoring and Evaluation
Design, refine, and audit reusable agent skills by creating structured SKILL.md files and evaluating trigger precision. The process includes auditing skill collections for redundancy and designing testable behaviour through evaluation promp…
Concise Text Processing Assistant
An instruction set for configuring an agent to perform concise summarisation, translation, and code review using plain text. It enforces strict constraints on response length and prohibits markdown formatting to ensure minimal overhead.
Integrating MDMA into Applications
This skill facilitates the integration of MDMA into applications by configuring parsing, runtime stores, React rendering, and LLM streaming. It provides guidance on leveraging the @mobile-reality package ecosystem for custom components, pro…
Claude Skill Design Guide
Provides a framework and best practices for designing modular, self-contained agentic skills that extend Claude's capabilities through specialised knowledge and tool integrations.
Swashbuckling Pirate Persona Skill
This skill transforms agent responses into a swashbuckling pirate persona using maritime idioms. It recontextualises tasks, problems, and data as quests, troubles at sea, and treasure.
Shakespearean Persona and Style Skill
This skill enables an agent to adopt the persona of a Shakespearean scholar, utilising early modern English and poetic cadence. It transforms model outputs into the distinctive style of William Shakespeare.