The "python" topic on GitHub groups 120 open-source projects in the DeepSeek Harness (DSH) ecosystem, led by notebooklm-py with 19.3k GitHub stars. notebooklm-py — Unofficial Python API and agentic skill for Google Gemini Notebook. Every project here is indexed by DSH Universe with live GitHub data — stars, activity and install status — so you can compare and install directly.
Unofficial Python API and agentic skill for Google Gemini Notebook. Full programmatic access to NotebookLM's features—including capabilities the web UI doesn't expose—via Python, CLI, and AI agents like Claude Code, Codex, and OpenClaw.
Autoharness — a self-learning skill layer for Claude Code — distills skills from your real sessions, updates them as you work, and prunes the ones that stop getting used. No daemon, no benchmark.
Local, private, open-source, self-evolving cross-platform AI content discovery Agent: understands you first, then proactively finds content from Bilibili, Xiaohongshu, Douyin, YouTube, X, Zhihu, Reddit, Weibo and the open Web.(supports deepseek harness plugin) | Local-first open-source cross-platform AI content discovery agent: understands you, then proactively finds content across Bilibili, Xiaohongshu, Douyin, YouTube, X, Zhihu, Reddit, Weibo and the open web.(support deepseek harness plugin)
An IOS Simulator Skill for ClaudeCode. Use it to optimise Claude's ability to build, run and interact with your apps, and to proxy xcodebuild to save token and context wastage.
Agent skill + renderer: turn one JSON config (or an Excalidraw file) into an animated architecture diagram — terminal-style live panel or pastel infographic — as mp4 or a looping web page. 动态架构图 / agent 架构动图,适配 X 与小红书。
[dsh] A more powerful visual toolkit for text-only models: one-line install and use, paste images for direct recognition, multi-image Q&A, screenshot-to-frontend UI restoration, and more|DeepSeek Harness-native integration for agent-vision-toolkit: image Q&A, long-screenshot OCR, UI restoration, grounding, pixel diff, Artifacts, and Web UI.
Local long-term memory for Hermes, Codex, Claude Code, WorkBuddy and dsh agents: one SQLite store, source-backed recall, and tools to inspect, correct or delete what is remembered.
📚 A zero-dependency, git-backed micro-lesson library for AI Agents to asynchronously share and search verified debugging experience. Python stdlib only. | https://misakanet.org
E-invoice organizing and reimbursement prep tool: batch-collect PDF/OFD/XML invoices from email, OCR, categorize and archive, and generate Excel summaries; offers Windows/macOS desktop versions and a DSH plugin.
Delegate bounded coding tasks to a headless WorkBuddy/CodeBuddy CLI agent from DeepSeek Harness — structured result, real diff, path whitelist and worktree isolation.
Draw the law · Make the Law Visible —— litigation visualization toolkit for legal professionals: redraw messy litigation diagrams into materials-ready diagrams, or directly read case materials and draw an accurate timeline. Works with Claude Skill / DeepSeek Harness.
One command. Full stack. Zero compromise. — All-in-one Claude Code skill with 33 modes, 6-layer security, 23 hooks, and 75% token savings. Works on Codex, Cursor, Manus, Windsurf.
🐋 DeepSeek Big Fat Fish desktop pet · standalone Windows desktop pet: walks around, checks the weather, and keeps an eye on your DeepSeek balance — balance / used today / peak and off-peak hours / per-turn consumption stats; drag snapping, left-snap flip, bouncy sound effects, number rolling, custom appearance, app integration. Desktop-pet version of the DeepSeek Harness (DSH) same little whale balance widget, MIT open source
DSH Director Toolkit is a DeepSeek Harness plugin for 3D artists, technical designers, and creative coders. Paste a half-formed idea, a reference note, or a portfolio caption and get a compact direction pack for Blender, Three.js, Houdini, or C4D.
Per-session environment selector for DeepSeek Harness: Python / R / CLI slots from conda, standalone R, WSL, or custom paths, via /env, session_env, DSH_ENV_* facts, and a header dropdown.
Resilient multi-provider web search & content extraction for Claude Code, DeepSeek Harness, and any MCP host — quota-aware failover, JS rendering, and a CLI.
Paper Plane X is a local-first workbench for research reading, paper processing and review writing. It chains PDF parsing, structured paper extraction, fact checking, project files, literature search and external Agent tools into a reusable research pipeline.
MCP server unifying Taobao / JD / Pinduoduo product search, detail and affiliate link conversion. Official open-platform APIs only, and it runs in mock mode before you have any credentials.
Football data MCP Server for Claude, Codex, Cursor and DeepSeek Harness, supporting 18 European leagues, 8 cups, plus match, standings, head-to-head, odds and recent form queries.
aisync unified entry for multiple computers: SSH reverse tunnel + nginx + Authelia aggregates the DeepSeek Harness Web GUI of several machines under one domain (includes ECS control plane / machine-side agent / one-click deploy script)
Durable agent runs for local fleets, built on the DeepSeek Harness (dsh). Streamlined Orchestra rewrite: one execution path, stdlib Python, SQLite, three-file console.
Gives the agent native browser control through browser-use: the package installs itself into a managed environment, updates itself on every launch, and answers through first-class tools.
Client-neutral AI-agent control plane: user global config overlay for Hermes/Codex/DSH/GitHub/Open Design/OpenHuman via one adapter contract (CC Switch observe-only). Modules: client-neutral-core (task/telemetry ledger, sidecar), services (orchestration/policy/receipts), read-only observer. Authority chain + exact-SHA CI gates.
One-click deploy DeepSeek Harness (dsh) to your server, local SSH tunnel ready out of the box | One-click DeepSeek Harness deployer with SSH tunnel access for Windows
Determine whether Python code is original or derivative work — pure local AST analysis, zero dependencies, offline. Agent Skill, usable by any agent. Detect derivative work in Python code, offline & dependency-free.
Pesquisa web para agentes de LLM via API Tavily com rotação AUTOMÁTICA de chaves — se a request morrer (429/432/401/5xx/rede), puxa outra chave e refaz a MESMA request; o agente nunca vê o erro. Funciona em qualquer agente/terminal, dentro ou fora do DSH.
Agent skill: 文件夹分类整理 + 可重建索引 —— index-first folder organization with a regenerable index (_INDEX.md/_index.json/_index.tsv) and a reference gate before any move.
Office file processing and token-saving workflow. Three unique conventions — Excel goes entirely through SQLite (import → SQL query → export report); Word drafts go Markdown → docx (pandoc) then fine typography; PPT auto-selects scenario layouts by purpose and enforces official-document review before generation. Interactive grouped questions by default, executes only after confirming the checklist.
Layered long-term memory for coding agents: decaying episodic recall (L2), a provenance-tracking semantic graph (L3), and budgeted procedural skills (L4) in one embedded SQLite file. Deterministic retrieval, no model calls, no server. Ships a DeepSeek Harness plugin.
A simulation of digital signal processing techniques for suppressing laser phase noise in fiber-optic sensing systems. The project compares raw signals, Wiener filtering, and reference delayed self-heterodyne (DSH) techniques to recover low-frequency vibration signals using affordable commercial lasers.
Local Pillow image processing MCP: 12 tools (info/resize/crop/convert/compress/rotate/flip/thumbnail/watermark/effects/placeholder/overlay), offline and zero-cost, connected to DSH via dsh-mcp-client. Local Pillow image MCP with 12 tools for DeepSeek Harness, offline & free. | Platforms: macOS/Windows/Linux (Python+Pillow)
An agent skill that batch-researches a YouTube channel's latest N videos — captions first, local ASR fallback — and writes one cross-video report instead of N summaries. Works in DSH, Codex, Claude Code, Hermes and WorkBuddy.
DSH bundle plugin: author Python tools in the Web UI or through agent tools, register them as first-class DSH tools, gate dangerous runs through built-in approval, group tools into packs.
DSHA ↔ Auto.js Pro bridge: read/tap nodes in apps DSHA cannot see (via AutoJsPro MCP) + drive AutoJsPro remotely to run any script (via its debug RPC).