Unified LLM API with automatic model discovery and provider configuration
Directory
Browse agent skills
Search by intent: design systems, testing, deployment, security, documentation, Azure, Supabase, React, or any workflow you want an agent to perform better.
Showing 16,861–16,890 of 127,293 skills
Page 563 of 4,244 · loading 30 at a time
AI SDK provider that drives a DeepSeek Harness (dsh) runtime as a language model (LanguageModelV3 — works on both AI SDK v6 and v7)
SSH remote-execution plugin for DeepSeek Harness — ProxyJump chain, SFTP filesystem, subprocess and PTY over ssh2
API client and device models for Papouch's devices
DeepSeek Harness Cordis tools for All in Flash, powered by All in Luna
dsh-tianshu-tui: an interactive TUI layer over the dsh-base profile, installable as a plugin (`dsh plugin --profile tui add @huiliyi37/dsh-tianshu-tui`)
Golden Frijoles Growth Engine TypeScript SDK for telemetry, experiments, and local feature-flag evaluation.
User-side undo/redo for DeepSeek Harness (dsh): the /undo command rolls the transcript back to before the last message and /redo restores it.
Local-first failure postmortems for DeepSeek Harness sessions.
Bounded multi-agent orchestration for the Pi coding agent: one writer, tool-layer governance, and verified acceptance.
Vue 3 renderer for Corpusctl headless CMS block content — zero default styling, every block type overridable
React renderer for Corpusctl headless CMS block content — zero default styling, every block type overridable
Typed SDK for Corpusctl headless CMS — edge-first ReadClient, ManagementClient and framework-agnostic block rendering (toRenderTree, renderToHtml)
A dairy-free RAG pipeline for delicious semantic similarity, clustering and NER.
Least privilege for AI coding agents: zones with owners, a five-state board with a gate, and grants bound to a task and a branch.
Automated tool-call approval for DeepSeek Harness: an `auto` tier for approval policy that classifies each tool call as allow / deny before dispatch
Agent skill that draws architecture, flowchart, sequence, data-flow and lifecycle diagrams as hand-placed SVG, to one linted house style.
Self-hosted document viewer: per-recipient tracked links, reading analytics, live presentation. The core knows nothing about the application hosting it — everything it borrows arrives through an injected context.
Bayesian network utilities shared by KBE and Ancestra Health: loopy belief propagation inference, diagnostic values, continuous state transitions and Noisy-MAX conversion.
Lightweight, developer-first AI prompt security middleware for real-time detection and remediation of sensitive data.
Turn any technical book PDF into a Claude Code skill — ready to study, reference, and use while you work.
RelationalAI Predictive Reasoner SDK
把 [grok-build](https://github.com/xai-org/grok-build) 的 TUI 作为 [DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness) (dsh) 的前端:界面是 grok 的,内核(提示词、工具、模型路由、会话持久化)由 dsh 提供。
@mdgate/pdf is a docs & writing skill indexed by Skillful.sh. Package: @mdgate/pdf.
Manual approval for Deepseek Harness (aka "Manual Mode"/"Ask Mode")
郁金香金融科技 - 工业级量化交易系统 (357策略 + RAG知识检索 + 多模型Fallback + API版本管理 + MT5实盘)
Shared `ctx.traits` CDK doctrine for reviewed multi-agent procedures, and the vendorable canonical `agents` trait package built from that same doctrine.
dsh-zcf: zero-config DeepSeek Harness setup wizard — install dsh, store provider credentials, and pick a runnable profile
Mini Program Engineering Skill Suite is an Agent Skill suite for evidence‑first mini‑program development. It helps agents bring WeChat or other mini‑program projects from vague intent to reliable engineering work: project intake, product specification, architecture, implementation, UI & device adaptation, debugging, verification, release readiness.