数据与 AI技能

Generalised autonomous optimisation loop — soft RLVR for any artifact a user can measure. Web runtime package: uses memory in this order: connector-backed, project-pack, none. Never assumes subprocess access or unrestricted local files. Use this skill whenever a user wants to iteratively improve an artifact — code, prompts, documents, configs, designs, content — by running structured experiments, evaluating results against a multi-dimensional rubric, and learning from each attempt. Triggers include: "optimise this", "keep improving until it's good", "run experiments on", "autoresearch", "iterate on this overnight", "try different approaches and pick the best", or any request implying repeated evaluate-and-improve cycles.

中文简介正在准备中。当前显示原文说明。

Boot and drive the tbh-meter app + reader locally to see a change working — what runs on macOS (UI only; the reader is win32-only) and how to feed the overlay/list fake data. Use when asked to run/start the meter, verify a change in the running app, screenshot the overlay, or reproduce a bug locally.

中文简介正在准备中。当前显示原文说明。

应用从 AI for Everyone 蒸馏出的「ML项目流程 — 收集→训练→部署→迭代」方法论。Use when the user needs practical guidance, diagnosis, workflow design, decision support, or an action plan related to ml-workflow.

中文简介正在准备中。当前显示原文说明。

应用从 AI for Everyone 蒸馏出的「统一数据仓库 — AI的数据基础设施」方法论。Use when the user needs practical guidance, diagnosis, workflow design, decision support, or an action plan related to unified-data-warehouse.

中文简介正在准备中。当前显示原文说明。

应用从 AI for Everyone 蒸馏出的「ML产出软件,DS产出洞察 — 项目类型判定框架」方法论。Use when the user needs practical guidance, diagnosis, workflow design, decision support, or an action plan related to ml-vs-ds.

中文简介正在准备中。当前显示原文说明。

应用从 AI for Everyone 蒸馏出的「不等完美数据 — 先用起来再迭代」方法论。Use when the user needs practical guidance, diagnosis, workflow design, decision support, or an action plan related to dont-wait-perfect-data.

中文简介正在准备中。当前显示原文说明。

当用户想系统评估一个AI项目是否值得投入时使用，结合概念简单性和数据充足性两个维度。 用户在问"这个想法可行吗"、"我们需要多少数据"、"技术上能做到吗"时激活。 不适用于: 纯数据科学洞察项目、探索性分析、已明确不需要AI的场景。

中文简介正在准备中。当前显示原文说明。

应用从 AI for Everyone 蒸馏出的「数据飞轮 — AI产品自我强化的增长引擎」方法论。Use when the user needs practical guidance, diagnosis, workflow design, decision support, or an action plan related to data-flywheel.

中文简介正在准备中。当前显示原文说明。

应用从 AI for Everyone 蒸馏出的「不为数据而收购 — 数据投资的价值验证原则」方法论。Use when the user needs practical guidance, diagnosis, workflow design, decision support, or an action plan related to dont-acquire-for-data.

中文简介正在准备中。当前显示原文说明。

应用从 AI for Everyone 蒸馏出的「统计化验收标准 — 不追求100%准确」方法论。Use when the user needs practical guidance, diagnosis, workflow design, decision support, or an action plan related to statistical-acceptance.

中文简介正在准备中。当前显示原文说明。

当用户想快速判断一项任务是否能被当前AI技术自动化时使用。 用户在问"这个AI能做吗"、"技术上可行吗"、"能不能用AI自动完成X"时激活。 不适用于: 需要深度可行性分析的项目评估、涉及情感共鸣或复杂推理的创意任务、LLM时代的长文本生成场景。

中文简介正在准备中。当前显示原文说明。

Writes and reviews Swift App Intents code that exposes app actions and data to Siri, Shortcuts, Spotlight, widgets, Control Center, and Apple Intelligence. Use when adding AppIntent, AppEntity, OpenIntent, AppShortcutsProvider, EntityQuery, Focus Filters, AssistantEntity/AssistantIntent schemas, on-screen awareness, LongRunningIntent/CancellableIntent, EntityCollection, SyncableEntity, AppIntentsTesting, or when wiring SwiftData/networked data into intents.

中文简介正在准备中。当前显示原文说明。

5つの専門エージェントがディスカッション型で競合のビジネスモデル・機能・価格・弱点を網羅的に分析し、 差別化ポイントと参入戦略を導出するスキル。 エージェント同士が対話・批判し合いながら調査し、ca-data-critic がライブモデレーターとして矛盾を検出・収束を判定する。 Phase 1 で競合プロファイル・レビュー・機能をディスカッション型で調査し、 Phase 2 でポジショニング戦略家が統合。 Gate 1（中間合意形成）→ Gate 2（最終検証 + Independent Validation + Devil's Advocate）の2段階ゲートで品質を担保。 Use when: 競合を分析したい、差別化ポイントを見つけたい、 市場のプレイヤーを調べたい、ポジショニングを考えたい。 Triggers: "競合分析", "competitive analysis", "競合調査", "差別化", "ポジショニング", "competitor", "競合", "市場プレイヤー", "ライバル調査", "competitive landscape"

中文简介正在准备中。当前显示原文说明。

Use when the user needs to design, run, or audit product or GTM analytics — metric hierarchies and north stars, funnels, cohort retention, churn, pipeline velocity, forecasting, lead scoring, NRR, A/B tests, or analytical SQL — or uploads a dataset to analyze. Triggers on questions like "why did activation drop", "is this metric meaningful", "review my dashboard", or "what should I measure". For writing the spec that defines a feature's success metrics, use product-spec.

中文简介正在准备中。当前显示原文说明。

Create or improve a Senpai target repository's program.md. Use this skill whenever the user wants to point Senpai at a fresh ML or research target repository, define the research objective, primary metric, benchmark contract, allowed edit boundaries, W&B reporting contract, or prepare a repo for autonomous advisor/student experiment loops.

中文简介正在准备中。当前显示原文说明。

Comprehensive primary skill for agents working with Weights & Biases. Covers both the W&B SDK (training runs, metrics, artifacts, sweeps) and the Weave SDK (GenAI traces, evaluations, scorers). Includes helper libraries, gotcha tables, and data analysis patterns. Use this skill whenever the user asks about W&B runs, Weave traces, evaluations, training metrics, loss curves, model comparisons, or any Weights & Biases data — even if they don't say "W&B" explicitly.

中文简介正在准备中。当前显示原文说明。

Analyze datasets and extract insights. Use when user needs to understand data patterns, statistics, or trends.

中文简介正在准备中。当前显示原文说明。

Use when turning perf, DTrace, pprof, or async-profiler stacks into an SVG flamegraph with the FlameGraph scripts, reading one, or diffing two profiles. Not for collecting data: use linux-perf.

中文简介正在准备中。当前显示原文说明。

Use when writing a GNU ld script for a bare-metal target, placing code in a flash or RAM region, wiring .data and .bss startup, or fixing a region overflowed error. Not for LTO: use linkers-lto.

中文简介正在准备中。当前显示原文说明。

Use when writing eBPF programs with libbpf, bpftrace, or Rust with aya-ebpf and aya-log, attaching kprobes, tracepoints, or XDP hooks, declaring maps, sharing data with a Tokio loader, triaging verifier or Aya load errors, or porting with CO-RE. Not for full kernel bypass: use dpdk.

中文简介正在准备中。当前显示原文说明。

Use when profiling with Intel VTune or AMD uProf for hotspots, top-down pipeline stalls, memory-bound analysis, or roofline data. Not for raw perf stat counters: use linux-perf.

中文简介正在准备中。当前显示原文说明。

Use when explaining pipeline stages, data or control hazards, forwarding, stalls, or superscalar basics behind a counter reading. Not for mispredict cost: use branch-prediction-and-speculation.

中文简介正在准备中。当前显示原文说明。

Use when writing reset-to-main startup code, vector tables, VTOR, .data/.bss init, stack setup, startup.s, or crt0 for Cortex-M/RISC-V. Not for the bootloader jump: use bootloaders-embedded.

中文简介正在准备中。当前显示原文说明。

Use when work reaches a session boundary or the user asks to hand off, delegate, or get a clipboard-ready prompt. Mode brief packages continuity; mode prompt builds the delegation prompt.

中文简介正在准备中。当前显示原文说明。