Art & Design

Codex Skill for DeFi contract discovery, economic-model analysis, fixed-cutoff data collection, and auditable reporting

OpenClaw UI theme coach: derive accessible CSS palettes, inject/restore themes, auto-heal snapshots

Codex CLI as the executor: Claude specs, reviews and verifies while Codex types. Five selectable engines — Luna/Sol/Terra on OpenAI, DeepSeek and GLM-5.3 Flash via Codex Router.

The DSH productivity suite: plugins + agent skills for the DeepSeek Harness, installable in one command.

Claude Code skill: bring Bose SoundTouch speakers back after the SoundTouch cloud shutdown - self-hosted replacement service, recovering each station's direct stream, and proving every preset plays.

Agent Skill for AI story writing: turns one prompt into a structured story using John Truby's Anatomy of Story. Blueprint then prose. Unofficial.

name: nmg-memory

Agent Skill and MCP backend for standalone or remote Jupyter Servers over HTTP and WebSocket

Agent Skill — audit agent definitions with a hybrid cost model: always-loaded descriptions as residency, bodies as invocation payload. AKC Curate-phase skill, third stocktake sibling.

Agent Skill: design LLM-as-judge evaluators with binary checks as evidence and one named holistic verdict — no score aggregation

Agent skill that publishes a finished skill as an independent GitHub product with its own story, install path, SEO/GEO metadata, and verification.

Universal agent skill for moving into and living in 1F3D9.

精简版的 Superpowers ，仅保留头脑风暴-计划-执行的核心步骤，去掉所有强制性流程骨架。

Agent skill that cleans up dictated, mixed-language, or self-correcting input while preserving the user's meaning and directness.

Agent skill that stress-tests SKILL.md files, system prompts, and agent rules across persona and scenario simulations before release.

Portable, versioned agentic-development skill library: lifecycle skills, a per-project adapter contract, and a cross-platform installer. Distilled from next-web-boilerplate.

🤖 Just a skill to let AI generate documents based on your source code 🔅

Agent skill that turns one subjective failure mode and real labeled traces into a binary Pass-Fail LLM evaluator prompt.

OpenClaw task & project progress tracking via PROGRESS.md cards (任务进度跟踪器：目录即容器，PROGRESS.md 即进度)

Your agent shipped. Aftertaste checks whether the UI still looks like everyone else's.

Audit and fix project documentation role overlaps — one command to keep CLAUDE.md, CODEMAPS, ADR, and README healthy

Beautiful visual thinking skills for AI agents — turn prompts into typed diagram sources, SVG, HTML and optional PNG with DiagramSpec.

Graph-based orchestration for coding agents. Run fewer agents. Verify every result.

Audit a project repo's non-code assets (tool configs, CI workflows, runbooks) for diminished value — Keep/Update/Retire/Merge. A semantic value layer above structural linters.