Production-grade skills for AI coding agents by Agents365-ai. Agent-agnostic, works with Claude Code, Cursor, Copilot, OpenClaw & more.
English | 中文
# Agent-agnostic — any agent (Claude Code, Cursor, Copilot, etc.)
npx skills add Agents365-ai/365-skills -g
# Claude Code plugin marketplace (optional)
/plugin marketplace add Agents365-ai/365-skillsInstall a single plugin (Claude Code): /plugin install <plugin-name>, e.g. /plugin install mermaid.
| Plugin | Description |
|---|---|
agent-native-design |
CLI design for AI agents — evaluate, design, and refactor CLIs to serve humans, agents, and orchestration systems simultaneously |
pi-durable-dev-rules |
Pi Durable developer reference — build applications and tools on the durable agent harness (@earendil-works/pi-durable): storage backends and restart recovery, TypeBox tools with structured output and nested calls, durable tasks with checkpoints and ownership, subagents and child tasks, documents and migrations, code mode, per-conversation environments, hook guards, compaction, and the normative Pico5 specification. Includes the pi-ai model layer and 19 recipes from the package's 32 examples |
pi-plugin-cc |
Drive the Pi coding agent from Claude Code — model-agnostic delegation and code review. /pi:review with structured findings, /pi:adversarial-review, /pi:rescue, /pi:parallel-rescue, racing/fallback across providers, incremental review |
tmux |
Run, drive, and monitor terminal programs from a coding agent: detached sessions at a fixed size, send-keys input injection, capture-pane polling with a race-free sentinel, format inspection, four completion-detection strategies, and the pitfalls that break naive tmux scripts |
| Plugin | Description |
|---|---|
mermaid |
Mermaid diagrams — text-based, GitHub-native, auto-layout |
excalidraw |
Excalidraw — hand-drawn whiteboard style |
plantuml |
PlantUML — UML, C4, sequence, class diagrams |
tldraw |
Tldraw — infinite canvas, sketch-style |
| Plugin | Description |
|---|---|
semanticscholar |
Semantic Scholar — academic paper search, citation graph, recommendations, BibTeX export |
paper-fetch |
Paper PDF downloader by DOI / title — 7-source fallback (Unpaywall, S2, arXiv, PMC, bioRxiv, publisher, Sci-Hub) with batch mode and idempotent retries |
scholar-deep-research |
End-to-end literature review pipeline — 8-phase script-driven workflow, 7 federated sources (OpenAlex, arXiv, Crossref, PubMed, DBLP, bioRxiv, Exa), cross-source dedup, dual-backend citation chasing, parallel deep-read fan-out, mandatory self-critique, cited reports across 5 archetypes |
asta |
Ai2 Asta MCP — Semantic Scholar academic graph over MCP (no Python). Intent-to-tool routing, safe fields defaults, citation traversal, snippet evidence retrieval, and DOI/arXiv/PMID via externalIds |
journal-abbrev |
Journal name abbreviation lookup — ISO 4 + MEDLINE, multi-source cascade (JabRef → AbbrevISO → NLM), BibTeX rewrite with --idempotency-key, atomic cache rebuild, agent-native JSON envelope with stable error codes and dry-run |
journal-if |
Journal impact factor (JCR IF) lookup — find a journal's IF by name, compare IF across journals, and rank publication venue quality, backed by a bundled journals_if.csv dataset |
target-prioritization |
Multi-source drug-target due-diligence — turn a ranked gene list (e.g. scRNA-seq DE output) into a per-gene dossier across UniProt, OpenTargets, and PubMed, plus a local cross-lineage DE scan, then re-rank by a configurable composite score (cross-lineage convergence + druggability + disease genetics + tractability + novelty). Disease-agnostic |
figshare |
Figshare v2 REST API — search public datasets/articles, batch-download files by ID/DOI/URL, and create, update, publish, or multi-part-upload to your own articles (large-file 3-step upload flow) |
zenodo |
Zenodo REST API — deposit, publish, version, and search research artifacts (datasets, software, papers) with a citable DOI; sandbox-first, bucket-API uploads, full metadata reference and end-to-end shell examples |
scientific-thinking-general |
Structured scientific reasoning meta-skill — separates fact / evidence / hypothesis, ranks competing explanations by support, calibrates conclusion language to evidence strength, and defines interpretation boundaries |
grant-thinking-general |
Grant proposal reasoning meta-skill — reviewer-aware logic, fundability framing (significance, innovation, feasibility), scope control, and section-by-section diagnosis before any NSFC/NIH-style proposal is written |
grant-thinking-cn-biology |
Chinese NSFC-focused biology grant reasoning — mechanism-centered scientific questions, reviewer-aware logic, innovation discipline, feasibility and scope control across funding levels (youth / general / key) |
| scientific-thinking-biology | Biology-specific adaptation of scientific-thinking-general — marker vs driver, experimental-system validity, an 8-level biological evidence hierarchy, alternative-explanation and confound checks (composition, batch, redundancy), and system-aware conclusion language |
| Plugin | Description |
|---|---|
obsidian-organizer |
Keep a large Obsidian vault tidy — file new notes into the best-fit folder and audit/reorganize existing structure, driven by a single source-of-truth map note (00_Index/Folder_Map.md) inside the vault. Link-safe by design (moves go through the obsidian CLI so wikilinks auto-repair), propose-then-confirm for bulk changes |
zotero-dev-rules |
Zotero developer reference — Web API v3 (read/write, file upload, syncing, streaming, OAuth), the desktop client's internal JavaScript API, plugin development (Zotero 7–10), translators, and CSL citation styles |
obsidian-dev-rules |
Obsidian developer reference — packaging of docs.obsidian.md for building/reviewing TypeScript plugins (Plugin lifecycle, Vault/Editor APIs, events, modals, views, CodeMirror 6) and CSS themes, plus community submission policies |
| Plugin | Description |
|---|---|
ttscn |
Multi-platform Chinese TTS text-to-speech — 14 backends (Edge/Doubao/CosyVoice/Qwen/StepFun/Zhipu/Azure/Tencent/Baidu/MiniMax/Xunfei/ElevenLabs/OpenAI/Google), agent-native CLI with JSON envelope, schema introspection, voice cloning, SSML, emotion, dialects, filterable HTML comparison page |
imagencn |
AI image generation via Alibaba Bailian, ByteDance Volcano Ark & Tencent Hunyuan — 23 models, Chinese text excellence, rich terminal UI, smart config |
videogencn |
AI video clip generation with Chinese video models — text-to-video, image-to-video, first/last-frame and reference-to-video across Bailian (Wan/PixVerse/Kling/Vidu), Jimeng (doubao-seedance), MiniMax (Hailuo), Hunyuan |
assetseeker |
Search free commercial-use creative assets — photos, illustrations, icons, video footage, music, sound effects, and fonts across Pexels, Unsplash, Pixabay, Iconify, Freesound, Google Fonts and more |
bangumi-frames |
Bilibili anime frame & character organizer — download a bangumi/UP video (or local file), extract scene-change keyframes, split scenery vs character crops, cluster by CCIP identity or pull one character via a reference folder; optional OCR+LaMa subtitle/watermark removal |
yt2bb |
Repurpose YouTube videos for Bilibili — download via yt-dlp, transcribe with whisper, generate bilingual (English-Chinese) SRT subtitles, hardcode them with ffmpeg |
Plugins here are edited in place: change the files under plugins/<name>/, bump that plugin's version in .claude-plugin/marketplace.json, and commit both together. Their original source repos are private or frozen, so this marketplace is the only distribution channel for them.
Two public skills are deliberately not distributed from here; they live in their own repos and are installed from there:
- Agents365-ai/drawio-skill:
npx skills add Agents365-ai/drawio-skill -g - Agents365-ai/video-podcast-maker:
npx skills add Agents365-ai/video-podcast-maker/skills/video-podcast-maker -g
If this skill set is helpful, consider supporting the author:
WeChat Pay |
Alipay |
Buy Me a Coffee |
Tip |
Agents365-ai
- Bilibili: https://space.bilibili.com/441831884
- GitHub: https://github.com/Agents365-ai