modelscope
- 27 skills
- 0 followers
- 6 hours ago last updated
- ▌ Ms Hub · modelscope bundleModelScope unified operations entrypoint. Covers model/dataset search, download, and upload; repository management; Studio deployment; MCP service search, deployment, and configuration; and Skills Center search, install, and publish. Use this skill whenever the user mentions ModelScope or any platform operation. Use ms-studio-deploy for complex Studio deployment workflows; see this skill's references for the expanded MCP and Skills Center details.
- ▌ Ms Studio Deploy · modelscope bundleDeploy a local project to a ModelScope Studio. Supports Gradio, Streamlit, Docker, and static website types. Covers creation, code sync, deployment, log monitoring, plaintext/secret variable management, and automatic diagnosis and repair. Use when the user mentions deploying to ModelScope, Studio, the ModelScope community, ModelScope, Studio deployment, Gradio deployment, Streamlit deployment, Docker deployment, FastAPI deployment, or static website deployment, or wants to publish a local app, web app, or API service to the cloud. Also applies when the user encounters a Studio build failure or runtime error and needs to inspect logs, wants to update an already-deployed Studio, or wants to manage a Studio's plaintext variables or secrets. Not applicable to: Hub repository management, model/dataset operations, MCP/skill management, model training, or model evaluation.
- ▌ Map Verl Config · modelscopeMap VERL training configuration to AgentJet configuration. Find VERL config in verl_default.yaml, check for existing mappings in config_auto_convertion_verl.jsonc, add new mappings to ajet_default.yaml and the conversion schema, and optionally add parameters to AgentJetJob.
- ▌ Skill Normalizer · modelscopeConvert skills in non-standard formats to the standard Agent Skills `SKILL.md` format. Validates YAML frontmatter (name, description, license, compatibility, metadata, allowed-tools), directory structure (SKILL.md, scripts/, references/, assets/), and best practices. Use when the user asks to normalize, validate, or fix a skill.
- ▌ Monitor With Tmux · modelscopeMonitor training progress by reading tmux content with exponential backoff intervals (30s, 1min, 2min, 4min, 8min, 16min), analyze logs when anomalies occur, and provide fix suggestions
- ▌ Write Swarm Client · modelscopeCreate an active, dataset-driven AgentJet swarm client. Write agent_roll.py and agent_run.py that iterate through a dataset, execute agent workflows, and compute rewards for reinforcement learning training with AgentJet Swarm.
- ▌ Swarm Configuration · modelscopeHow `max_env_worker` caps the "Running Episodes" gauge, and how `AgentJetJob` relates to the YAML config.
- ▌ Train Complex Blackbox · modelscopeTrain complex blackbox agents (agents without clear reward signals) using AgentJet. Write dataset collectors, episode runners with LLM-as-Judge reward functions, and integrate with the AgentJet training loop.
- ▌ Install Agentjet Client · modelscopeInstall AgentJet client for connecting to a swarm server. Use when the user only needs to run the AgentJet client (not a swarm server) and does not need to run models locally, e.g. on a laptop. Installs basic requirements via `pip install -e .`.
- ▌ Download From Swanlab Url · modelscopeDownload per-step time-series metric data (reward, entropy, response length, etc.) from a SwanLab cloud run URL as a pandas.DataFrame. Use when the user provides a SwanLab URL and wants to fetch or analyze training curves.
- ▌ Write Passive Swarm Client · modelscopeCreate a passive swarm client that waits for user input instead of iterating through a dataset by itself.
- ▌ Uv Install Agentjet Swarm Server · modelscopeInstall AgentJet swarm server using the UV package manager. Handles virtual environment creation with Python 3.10, dependency installation with the verl training backbone, flash-attn compilation, and optional PyPI mirror for China users.
- ▌ Conda Install Agentjet Swarm Server · modelscopeInstall AgentJet swarm server using Conda. Handles Python 3.10 environment creation, dependency installation with the verl training backbone, flash-attn compilation, and optional PyPI mirror for China users.
- ▌ Docker Install Agentjet Swarm Server · modelscopeInstall and run the AgentJet Swarm Server in a Docker container with NVIDIA GPU support. Use when the user wants to deploy a swarm server on a GPU machine via Docker, including GPU driver setup, Docker mirror configuration, model weight mounting, and server startup.
- ▌
- ▌ Create Keep Think Model Chat Template · modelscopeYour task is to investigate the chat template of given model, go to its tokenizer config and check whether the following behavior exists:
- ▌
- ▌ Evalscope · modelscope bundleLLM evaluation & inference performance testing via the evalscope CLI. Translates natural language requests into evalscope commands for: (1) Model accuracy evaluation — runs registered benchmarks against local checkpoints or API endpoints (OpenAI-compatible, Anthropic, LiteLLM); (2) Performance stress testing — TTFT, TPOT, throughput, latency under configurable concurrency; (3) RAG evaluation — RAGAS quality metrics, MTEB embedding benchmarks, CLIP retrieval; (4) Benchmark discovery — list/filter/inspect benchmarks by tag. Trigger on: evaluate / benchmark / score a model, throughput / latency / QPS / stress test, find benchmarks, view results, 评测模型, 压测, 跑 benchmark, 性能测试, 查看评测结果, 有哪些评测集, RAG 评测, embedding 评测. Do NOT trigger for: model training / finetuning / deployment / serving requests.
- ▌ Sirchmunk · modelscope bundleLocal file search using sirchmunk API. Use when you need to search for files or content by asking natural language questions.
- ▌ Ultron · modelscope bundleUltron: Collective Memory Synchronization System for General-Purpose Agents
- ▌ X Components · modelscope bundleUse when building AI chat UIs with @ant-design/x components — covers Bubble, Sender, Conversations, Prompts, ThoughtChain, Actions, Welcome, Attachments, Sources, Suggestion, Think, FileCard, CodeHighlighter, Mermaid, Folder, XProvider, and Notification.
- ▌ X Chat Provider · modelscope bundleFocus on implementing custom Chat Provider, helping to adapt any streaming interface to Ant Design X standard format
- ▌ Ms Agent · modelscope bundleAccess ms-agent's advanced AI capabilities via MCP tools: deep research, document research, financial research, code generation, video generation, web search (arxiv/exa/serpapi), LSP code validation (TypeScript/Python/Java), concurrent-safe file editing, and agent delegation. All project-level capabilities support async submit/check/get patterns. Use when the user asks to research a topic, analyze documents, generate code or videos, validate code, edit files, or delegate tasks. Requires ms-agent (pip install ms-agent).
- ▌ X Card · modelscope bundleUse when building AI-driven UIs with @ant-design/x-card — covers XCard.Box, XCard.Card, A2UI v0.9 commands, data binding, catalogs, actions, and streaming patterns.
- ▌ X Request · modelscope bundleFocus on explaining the practical configuration and usage of XRequest, providing accurate configuration instructions based on official documentation
- ▌ Use X Chat · modelscope bundleFocus on explaining how to use the useXChat Hook, including custom Provider integration, message management, error handling, multi-conversation management, and more
- ▌ X Markdown · modelscope bundleUse when building or reviewing Markdown rendering with @ant-design/x-markdown, including streaming Markdown, custom component mapping, plugins, themes, and chat-oriented rich content.