Total 52,846 skills, AI & Machine Learning has 8861 skills
Showing 12 of 8861 skills
Create a new Harbor task for evaluating agents. Use when the user wants to scaffold, build, or design a new task, benchmark problem, or eval. Guides through instruction writing, environment setup, verifier design (pytest vs Reward Kit vs custom), and solution scripting.
Explore and investigate ideas before committing to a change. Trigger: When the orchestrator launches you to think through a feature, investigate the codebase, or clarify requirements.
Complete guide to deploying OpenClaw on Alibaba Cloud servers, covering Alibaba Cloud Bailian model configuration, full-platform access for Feishu/DingTalk/QQ/Discord, Qoder CLI integration, as well as advanced configuration and O&M.
Detect whether an API endpoint is backed by genuine Claude (not a wrapper, proxy, or impersonator) using 9 weighted rule-based checks that mirror the claude-verify project. Also extracts injected system prompts from providers that override Claude's identity. Fully self-contained — copy the code below and run, no extra packages beyond httpx. Use when the user wants to verify a Claude API key or endpoint, check if a third-party Claude service is authentic, audit API providers for Claude authenticity, test multiple models in parallel, or discover what system prompt a provider has injected.
Bootstrap a local AI review pipeline and generate a paste-ready review prompt for any provider (Codex, Gemini, GPT, Claude, etc.). Use after creating a handoff or when ready to get an AI code review.
Simulate a senior high school grade 3 science tutor in China, using progressive teaching methods to tutor math, physics, chemistry, biology and other science problems. Used when students ask science questions, request explanations, say "I don't understand" or "Teach me". Suitable for Gaokao preparation, problem-solving guidance, and concept understanding.
Metric-learning recognition (ml-recog) for fine-grained visual recognition. Learns embeddings for retrieval-based matching (e.g., retail product recognition) using triplet / contrastive losses. Use when training, evaluating, exporting, or running inference for a TAO metric-learning recognition model. Trigger phrases include "train metric learning", "ml-recog", "retrieval embeddings", "triplet loss recognition", "fine-grained matching".
Runs the DEFT embed-then-mine workflow for VCN AOI iterations — embeds the gap-analysis target parquet, embeds a source pool, and mines nearest-neighbour source images for downstream augmentation. Use as the immediate next step after `tao-route-visual-changenet-samples` when expanding a real-image augmentation queue from the mining subset.
Orchestrate multi-phase deep research with web search, memory retrieval, pattern matching, and synthesis into structured findings
Generate a complete professional event plan from guest background material. Takes guest bio / CV / intro copy as input, runs a multi-turn clarification dialog (activity type, audience, duration, tone), and outputs a cohesive plan with title + promo copy, minute-level rundown, tiered host question set, and gift / takeaway suggestions. Trigger when user mentions "活动策划", "策划案", "嘉宾对谈", "沙龙策划", "主持人问题", "活动流程", "event planning", "host prep", "salon plan", or pastes a guest bio asking for an event plan.
Generate images using Gemini via ZenMux
Volcengine Doubao Search API (formerly Web Search/Fused Information Search), returns web page/image results. Prioritize using this skill for web search scenarios. Trigger words include: Doubao Search, check/search/find, is it true/reliable/confirm/verify, recent/today/latest/recent period, source/origin/link, Agent Plan, Fused Information Search, Harness, what are there/recommendations, price/policy/exchange rate/market trend, comparison/difference/which is better, I heard/said/not sure, hot search/popular/trending, help me look/understand, verify/rumor refutation, is it worth it/should I. Prioritize use when tasks rely on online facts or timeliness. If the answer may depend on external facts, call this skill first before answering.