DeepSeek agent skills, page 4

Guidance for using the MMSP TypeScript SDK (@prismshadow/mmsp).

Prism-Shadow/model-message-stream-protocol

Draw Sebastian-Raschka-gallery-style TikZ architecture diagrams for any HuggingFace decoder-only LLM, with per-block parameter formulas and concrete numbers.

yzlnew/infra-skills

Provides guidance for enterprise-grade RL training using miles, a production-ready fork of slime.

Orchestra-Research/AI-Research-SKILLs

Train Mixture of Experts (MoE) models using DeepSpeed or HuggingFace.

Orchestra-Research/AI-Research-SKILLs

Checks audit reports (financial statements and notes) for cross-reference, summation and text errors, with arithmetic done by script and results exported to Excel and Markdown.

nigo81/nigo-skills

Start the Claude Paper web viewer to browse and study papers

alaliqing/claude-paper

Run a one-shot or flat parallel batch of provider LLM subagents (headless cc-fleet subagent) that return a result.

ethanhq/cc-fleet

Chat / code generation via KeiRouter using OpenAI /v1/chat/completions or Anthropic /v1/messages format with streaming + auto-fallback combos.

mydisha/keirouter

Audit the repository or a broad subsystem for security, correctness, and maintainability risks.

ccch1mneyyy/dsh-TUI

Sets up and troubleshoots MCP API key authentication for a stock data service, covering key creation, client configuration and per-tool usage statistics.

Yourdaylight/stock_datasource

Minimize token consumption & maximize prompt cache hit rate.

lokikill123/codex-token-skills

Use TypeSafe Jev for narrow judgments inside DeepSeek Harness — routing, classifying, scoring, verifying, reranking — instead of spending a model turn on them.

PerryLink/jevcore

The user asks about videos watched in the past or across sessions — "have we watched anything about X", "which video showed that error", "what did that meeting decide", "search my videos", or a…

oxbshw/watch-skill

Review a provider/agent adapter implementation or PR — transcript discovery, parsing, and indexing for agent hosts (Claude, Codex, Kimi, Pi, DeepSeek Harness, OMP, Copilot, etc.).

tommy0103/obelisk

Describes ClawRouter, a local proxy that forwards each LLM request to the blockrun.ai gateway, which routes to a cheaper capable model, paid by USDC wallet or API key credit.

BlockRunAI/ClawRouter

Connect and use a user-owned Composio MCP server in Cetus. An agent skill from drewnekota/cetus.

drewnekota/cetus

Looks up kanban boards and task status through two bash-invoked tools instead of guessing, listing boards first and then filtering tasks by status, board or assistant.

YaoApp/yao

How to dispatch well — choosing flash vs pro, writing self-contained briefs, parallelism, verifying results, and guardrails

ZSeven-W/dsh-crew

Control a real browser via CDP — navigate, click, type, screenshot, and extract web content.

AgenticMatrix/coderix

Has several open-weight models answer a question, rank each other's anonymized answers, then lets a chairman model write the final response through Fireworks AI.

dair-ai/dair-academy-plugins

A skill your agent uses when a deepseek-harness change breaks an externally perceptible surface (CLI, profiles, cordis.yml or settings keys, persisted user data, SDK or wire APIs, published package…

Zhou-Yujing114514/deepseek-harness-linux

The user wants to share or hand off what was found in a video — "send this analysis to my team", "make a page I can share", "export the findings", "give me something I can attach to the ticket".

oxbshw/watch-skill

网文创作工具与生产流程。当用户需要使用AI辅助写作工具、设计提示词模板、规划写作工作流、选择码字工具、制定发布与数据分析策略时触发。覆盖DeepSeek/ChatGPT/Claude/豆包等AI工具应用、提示词工程、工作流设计、发布运营。

uu201/character-arc

Orchestrates the full WeChat official account workflow from topic to draft, chaining sub-skills for writing, review, layout, images and publishing.

aiworkskills/wechat-article-skills

Estimate GPU memory usage for Megatron-based MoE (Mixture of Experts) and dense models.

yzlnew/infra-skills

A skill your agent uses when writing or reviewing tests for DeepSeek Harness plugins, external DSH plugin packages, or package changes in the deepseek-harness repository.

NanmiCoder/dsh-auto-mode

把本周零散的工作记录整理成固定格式的周报。用户要求“写周报”、“整理本周工作”、“周报”时使用. An agent skill from Prism-Shadow/deepseek-harness-book.

Prism-Shadow/deepseek-harness-book

A skill your agent uses for xskill. An agent skill from SkillNerds/xskill.

SkillNerds/xskill

Three rules for multi-step work with DeepSeek V4 thinking models: verify references, use a verifier subagent before big edits, and write plans with exact path and line.

codewhale-hq/Codewhale

Turn a reported defect into an actionable bug report or issue draft.

ccch1mneyyy/dsh-TUI

Add new AI models to Kiln's mlmodellist.py and produce a Discord announcement.

Kiln-AI/Kiln

Select and run the smallest sufficient validation set for Grix changes before push, review handoff, or completion.

askie/grix

Queries historical A-share, Hong Kong stock, ETF and index data through an MCP server: daily K-lines, financial statements, market indicators and screening.

Yourdaylight/stock_datasource

The user shared a video URL, a YouTube/TikTok/stream link, a local video file, a screen recording, a meeting recording, or a playlist/folder of videos — "watch this", "summarize this video", "what's…

oxbshw/watch-skill

Fixed workflow for developing MMSP itself — adding or updating model support, and changing its pages.

Prism-Shadow/model-message-stream-protocol

Sends chat completions, streams responses, and opens an interactive REPL against any OmniRoute-routed model provider.

diegosouzapw/OmniRoute

Update this repository's NuGet package version and release notes, then prepare the changes for merge into main so CI can publish the release.

niltor/DeepSeekSDK-NET

Orchestrates repo pipeline from a single URL: branch Zhihu vs generic page → spider → DeepSeek caption when needed → Edge TTS → Remotion render to output/video/video.mp4.

szhshp/panda-video-generator

A skill your agent uses when the user hands over a dsh plugin repository (or a real migration commit / version corridor) and wants its upgrade experience extracted into one auto-graded Harbor…

NanmiCoder/dsh-auto-mode

A skill your agent uses for live web research, webpage fetching, login-state browsing, dynamic pages, social sites, and tasks that need current internet evidence.

itwanger/PaiCLI-Python

Breaks LLM torch profiler traces down by forward pass, layer and kernel, with timing tables and Perfetto time ranges for the layers you want to inspect.

BBuf/AI-Infra-Auto-Driven-SKILLS

Coordinate local worker agents (Grok Build, Kimi Code, Antigravity, DeepSeek Harness, OpenCode, Claude Code, Codex CLI, Devin CLI, ZCode, MiniMax Code) through the Agent Bridge MCP tools.

FeiZhuLulu/Agent-Bridge

Read and triage existing pull request review comments, or address them when requested.

ccch1mneyyy/dsh-TUI

Set up and use CALL-E for real outbound phone calls through the CALL-E CLI and MCP tools.

drewnekota/cetus

Set up the minimal set of artifacts (tokenized DCLM corpus shard + released HuggingFace checkpoint converted to DCP) required to benchmark, profile, or regression-test a MoE model in PithTrain.

mlc-ai/pith-train

Runs a brand's AI-search visibility work end to end: diagnosing how AI platforms represent it, repositioning it, producing AI-optimized content and monitoring ongoing mentions.

LeoYeAI/openclaw-master-skills

Persistent global memory across conversations. An agent skill from lokikill123/codex-token-skills.

lokikill123/codex-token-skills

新增 LLM 模型专家，指导如何在本项目中添加新的 LLM 供应商和模型。当需要接入新的大语言模型（如 OpenAI、Codex、DeepSeek 等）时使用。

jeffstric/ZJT