image-generation drafts prompts for dall-e, midjourney, and stable diffusion by optimizing keywords, style modifiers, and negative prompts across platforms and aspect ratios.

tldr-router maps natural language questions to tldr subcommands for code exploration. covers file discovery, structure inspection, and call graph traversal without prescriptive routing logic.

diffity-tour generates narrated code walkthroughs in the browser, responding to natural language questions or github pr urls. covers authentication, architecture, and feature explanations with sidebar navigation.

qwen-image-2-pro wraps alibaba's image model via inference.sh cli for text-heavy designs. covers basic invocation but lacks depth on parameters, constraints, and failure modes.

react documents styling conventions and component hierarchy for lobehub projects, prioritizing zero-runtime static styles over dynamic approaches and internal ui libraries over antd.

wandb-primary attempts to cover w&b project workflows, but the skill definition is incomplete. only environment defaults and a recipe section stub exist, with no actual procedures or decision logic. substantial content is missing.

minecraftconsoles-lce is a c++ reimplementation of minecraft legacy console edition targeting windows with lan multiplayer and dedicated server support, but the documentation conflates a game engine project with an ai skill.

react-native-dev is a reference index for expo and react native patterns. it maps to external docs but contains no executable procedures, decision logic, or actionable workflows for agents.

remotion-maps covers three map animation techniques for remotion projects: static satellite overlays, mapbox integration with globe support and 3d buildings, and maplibre as an alternative. technique selection is required upfront.

proactive-agent describes a concept for anticipatory ai agents with memory and self-improvement, but lacks procedural specificity, concrete implementation details, or actionable workflows.

fastapi-microservices-serverless outlines architectural principles for cloud-native fastapi services but lacks concrete procedures, actionable workflows, or decision logic to guide implementation.

ui-design outlines visual and interaction principles for accessible interfaces, but lacks actionable procedures, decision logic, or success metrics tied to measurable outcomes.

video-ad-analysis provides a framework for deconstructing ad creatives across hooks, persuasion tactics, and targeting patterns, but lacks executable steps, input/output contracts, and failure mode handling.

sf-ai-agentforce defines scope boundaries for declarative agent development via salesforce setup ui, covering metadata types and builder workflows. routing guidance is provided but implementation procedures are absent.

scaffolds python project structures with modern tooling including uv, fastapi, and django. generates type-safe applications with testing and configuration, though execution details remain vague.

translation-expertise outlines principles for english-japanese-chinese translation work but lacks actionable procedure, concrete triggers, or recovery paths. reads as philosophy rather than operational skill.

frontend-state-management catalogs patterns and libraries for managing application state, covering redux, zustand, context api, and mobx with boilerplate examples but minimal actionable guidance.

task-observer codifies ad-hoc skill friction capture during active work, treating continuous noticing as a formal process. declares intent clearly but lacks actionable procedure.

long-context documents transformer positional encoding techniques (rope, yarn, alibi, interpolation) for extending model context windows. lacks actionable procedures and failure mode discussion.

persona-researcher wraps google workspace apis to organize research artifacts across drive, docs, and sheets. lacks procedural depth and concrete trigger definitions.

cargo-workspace-management covers workspace administration through CLI including user/token/folder/role management and reporting, but lacks concrete procedural steps and actionable examples embedded in the skill itself.

agent-development covers design and implementation of custom claude code agents, emphasizing effective descriptions for task delegation, granular tool access control, and prompt-driven behavior encoding. guidance spans description patterns, memory tuning, and specification matching across task types.

meta-cognition-parallel attempts three-layer parallel analysis of rust design questions (language, patterns, domain) but lacks concrete procedures, failure modes, or execution details.

discover-brand coordinates multi-platform search for brand artifacts, triages results into a structured report, and optionally generates guidelines. vague on integration mechanics and error handling.

mm2-roblox-analytics-tracker claims to deliver inventory tracking and performance analytics for murder mystery 2, but lacks procedural depth, trigger specificity, and concrete implementation details.

etf-premium calculates whether an etf trades at a premium or discount to nav using yahoo finance data. designed for research and education, not live trading.

blogwatcher monitors rss and atom feeds via cli, offering basic add/remove/scan operations for tracking blog updates. lacks automation context and error handling guidance.

quickstart detects local environment setup (git, ao cli, agents directories) and provides initial orientation for new agentops users, but incomplete procedure and missing recovery paths limit utility.

threejs-lighting covers basic light type instantiation and scene attachment, but lacks procedural depth for shadows, ibl setup, performance tuning, or configuration workflows.

memory-management wraps an agentdb vector search backend with hnsw indexing for pattern retrieval and semantic lookup across distributed agents, but lacks procedural detail and failure handling.

masonry covers layout selection and comparison for varying-height content, distinguishing when masonry beats grid or card layouts for galleries and portfolios.

arize-prompt-optimization documents a four-phase workflow for extracting prompts from trace data, analyzing performance signals, and iterating on revisions using an llm-as-judge meta-prompt. heavy on jq and ax command reference, light on decision-making guidance.

zero-trust-architecture outlines identity-centric security principles and microsegmentation patterns for cloud-native systems, with incomplete code examples and generic best practice checklists.

product-photography generates studio-style product images via the belt CLI and bytedance seedream model. covers packshot fundamentals through prompt engineering but lacks procedural depth and failure handling.

bigquery-basics covers dataset and table creation, api enablement, and basic sql query execution via the bq cli tool. reference directory points to broader topics without procedural depth.

backtesting-frameworks outlines when to apply backtesting infrastructure but lacks actionable procedure. defines scope clearly with use/no-use cases, yet provides no numbered steps, concrete examples, or implementation guidance.

quarkus skill describes framework expertise and coding conventions but lacks executable procedures, decision logic, or concrete guidance on specific tasks like building native images or setting up reactive endpoints.

get-shit-done is a command dispatcher for structured project workflow, but lacks concrete procedure steps, failure handling, and actionable guidance on phase execution or verification.

ecosystem-primer maps the langchain/langgraph/deep agents technology stack and guides selection between abstraction layers based on execution durability and control flow needs.

analyze-project provides read-only exploration of deep learning repositories by mapping model structure, training and inference entrypoints, config relationships, and patterns. guidance is conservative, non-prescriptive.

context-master documents windows file path conventions (backslash vs forward slash) but lacks coherent skill definition, actionable procedure, trigger conditions, or outcome signals.

qdrant-scaling categorizes scaling problems into data volume, throughput, latency, and query volume, then defers to external docs without providing actionable guidance, decision logic, or concrete procedures.

shipping-products collects product leadership frameworks for accelerating release cycles. offers diagnostic patterns for identifying blockers and managing quality tradeoffs, but lacks concrete workflow automation.

alicloud-ai-image-qwen-image-edit-test is a minimal smoke test that validates the core request path for qwen image editing without specifying concrete inputs, outputs, or failure recovery mechanisms.

subtitle-generation catalogs caption and transcript capabilities (auto-transcription, multi-language, animation, export formats, speaker labeling) without actionable procedure, triggering logic, or failure handling.

pdf provides skeleton documentation for common operations (read, write, merge, extract) but lacks procedural depth, trigger patterns, and failure mode handling needed for reliable automation.

skill-security-auditor scans skills for security risks and produces pass/warn/fail verdicts. lacks detailed procedural steps, edge case handling, and specific trigger definitions.