Skills library

发布时间：2025年06月18日

# # 通过基于LLM的推理与行动智能体，实现复杂故障分析工作流的高效管理

发布时间：2025年06月18日

`Agent` `失效分析` `工业工程`

> Managing Complex Failure Analysis Workflows with LLM-based Reasoning and Acting Agents

# 摘要

> 失效分析（FA）是一项复杂且高度依赖专业知识的工作。在FA实验室中，通过引入AI组件，可以实现图像检测、案例检索和报告生成等任务的自动化。然而，随着AI模型数量的增加，如何将这些AI组件高效整合到FA工作…

--- id: a8d97bd4c764e91f9d2bda01 title: Binary Agents challengeType: 1 forumTopicId: 14273 dashedName: binary-agents ---

---
id: a8d97bd4c764e91f9d2bda01
title: Binary Agents
challengeType: 1
forumTopicId: 14273
dashedName: binary-agents
---

# --description--

Return an English translated sentence of the passed binary string.

The binary string will be space separated.

# --hints--

`binaryAgent("…

--- id: a8d97bd4c764e91f9d2bda01 title: Binary Agents challengeType: 1 forumTopicId: 14273 dashedName: binary-agents ---

---
id: a8d97bd4c764e91f9d2bda01
title: Binary Agents
challengeType: 1
forumTopicId: 14273
dashedName: binary-agents
---

# --description--

Return an English translated sentence of the passed binary string.

The binary string will be space separated.

# --hints--

`binaryAgent("…

--- id: a8d97bd4c764e91f9d2bda01 title: Binary Agents challengeType: 1 forumTopicId: 14273 dashedName: binary-agents ---

---
id: a8d97bd4c764e91f9d2bda01
title: Binary Agents
challengeType: 1
forumTopicId: 14273
dashedName: binary-agents
---

# --description--

Return an English translated sentence of the passed binary string.

The binary string will be space separated.

# --hints--

`binaryAgent("…

--- id: a8d97bd4c764e91f9d2bda01 title: Binary Agents challengeType: 1 forumTopicId: 14273 dashedName: binary-agents ---

---
id: a8d97bd4c764e91f9d2bda01
title: Binary Agents
challengeType: 1
forumTopicId: 14273
dashedName: binary-agents
---

# --description--

Return an English translated sentence of the passed binary string.

The binary string will be space separated.

# --hints--

`binaryAgent("…

--- id: a8d97bd4c764e91f9d2bda01 title: Binary Agents challengeType: 1 forumTopicId: 14273 dashedName: binary-agents ---

---
id: a8d97bd4c764e91f9d2bda01
title: Binary Agents
challengeType: 1
forumTopicId: 14273
dashedName: binary-agents
---

# --description--

Return an English translated sentence of the passed binary string.

The binary string will be space separated.

# --hints--

`binaryAgent("…

--- id: a8d97bd4c764e91f9d2bda01 title: Binary Agents challengeType: 1 forumTopicId: 14273 dashedName: binary-agents ---

---
id: a8d97bd4c764e91f9d2bda01
title: Binary Agents
challengeType: 1
forumTopicId: 14273
dashedName: binary-agents
---

# --description--

Return an English translated sentence of the passed binary string.

The binary string will be space separated.

# --hints--

`binaryAgent("…

Find and install specialist modes for your project.

# /agents

Find and install specialist modes for your project.

user-invocable: true
description: Without arguments, scans your project and recommends specialist modes. With a search query, searches for specific modes. Usage: /agents [search query]
allowed-tools: [Read, Glob, Gre…

--- title: "With Flink Agents" url: "https://flink.apache.org/getting-started/with-flink-agents/" date: "Mon, 01 Jan 0001 00:00:00 +0000" author: "" feed_url: "https://flink.apache.org/index.xml" --- Getting Started with Flink Agents # Read how you can get started with Flink Agents here.

---
title: "With Flink Agents"
url: "https://flink.apache.org/getting-started/with-flink-agents/"
date: "Mon, 01 Jan 0001 00:00:00 +0000"
author: ""
feed_url: "https://flink.apache.org/index.xml"
---
Getting Started with Flink Agents # Read how you can get started with Flink Agen…

         

![](SS/SS-10272025_20.png)
![](SS/SS-10272025_21.png)
![](SS/SS-10272025_22.png)
![](SS/SS-10272025_23.png)
![](SS/SS-10272025_24.png)
![](SS/SS-10272025_25.png)
![](SS/SS-10272025_26.png)
![](SS/SS-10272025_27.png)
![](SS/SS-10272025_28.png)
![](SS/SS-10272025_29.png)

--- question: 哪些 coding agent 支持 SKILL.md？ order: 5 ---

---
question: 哪些 coding agent 支持 SKILL.md？
order: 5
---

Skills 适用于支持 `SKILL.md` 格式的 coding agent。注册表跟踪已测试兼容性的常见 Agent，包括：

- **Claude Code**
- **Claude API**
- **Codex**
- **Cursor**
- **GitHub Copilot**
- **Google Gemini**
- **Kiro**
- **OpenCode**
- 以及更多

> **Como usar e criar providers de IA no FinEngine. Sistema 100% plugável — sem lock-in.**

# 06 — AI Agents

> **Como usar e criar providers de IA no FinEngine. Sistema 100% plugável — sem lock-in.**

**Navegação:** [← Core Engine](05-core-engine.md) | [Database →](07-database-supabase.md)

---

## Índice

- [Visão geral](#visão-geral)
- [Como funciona](#como-funciona)…

--- doc_type: short full_text: sources/0683 - GardenDesigner - Encoding Aesthetic Principles into Jiangnan Garden Construction via a Chain of Agents.md ---

---
doc_type: short
full_text: sources/0683 - GardenDesigner - Encoding Aesthetic Principles into Jiangnan Garden Construction via a Chain of Agents.md
---

# 0683. GardenDesigner: Encoding Aesthetic Principles into Jiangnan Garden Construction via a Chain of Agents

**Source typ…

- Use TDD with `pytest` - Use `uv`

## General workflow

- Use TDD with `pytest`
- Use `uv`

## Issue Tracking with bd (beads)

**IMPORTANT**: This project uses **bd (beads)** for ALL issue tracking. Do NOT use markdown TODOs, task lists, or other tracking methods.

### Why bd?

- Dependency-aware: Track blockers a…

Definition and Core Concepts

# 1.1 What are AI Agents?

## Definition and Core Concepts

An artificial intelligence (AI) agent is a software program that can interact with its environment, collect data, and use that data to perform self-determined tasks to meet predetermined goals. Unlike traditional AI syst…

Llegados al capítulo 10, el harness tiene costuras limpias por todos lados: el provider (capítulo 03), la compactación (capítulo 07), las herramientas (capítulo 09) y los paquetes `internal/` que le dan un sitio a cada capa. Pero todavía no sabe *delegar*. Toda llamada a herramienta acaba volviendo a la conversación principal. Leer veinte archivos para responder a una sola pregunta llena el contexto con veinte bloques tool-result y el bucle principal ya no recupera el hilo. Este capítulo introduce la abstracción que lo soluciona.

# 11 · Subagentes

Llegados al capítulo 10, el harness tiene costuras limpias por todos lados: el provider (capítulo 03), la compactación (capítulo 07), las herramientas (capítulo 09) y los paquetes `internal/` que le dan un sitio a cada capa. Pero todavía no sabe *delegar*. Toda…

--- title: 92% of Security Pros Concerned About AI Agents url: https://www.darktrace.com/blog/state-of-ai-cybersecurity-2026-92-of-security-professionals-concerned-about-the-impact-of-ai-agents date: '2026-03-31' author: '' feed_url: https://darktrace.com/blog/rss.xml --- The State of AI Cybersecurity 2026 report reveals growing concern among security professionals as AI adoption accelerates, highlighting the urgent need for scalable strategies to secure AI systems.

---
title: 92% of Security Pros Concerned About AI Agents
url: https://www.darktrace.com/blog/state-of-ai-cybersecurity-2026-92-of-security-professionals-concerned-about-the-impact-of-ai-agents
date: '2026-03-31'
author: ''
feed_url: https://darktrace.com/blog/rss.xml
---
The Sta…

--- title: "【技術解析】讓電腦使用 Agent 不再昂貴：Yale NLP 的台階層級效率優化框架" description: "Yale 與 UNC 研究者提出 event-driven 級聯架構，讓 AI Agent 在簡單步驟用小模型、風險時刻才叫用大模型，在 OSWorld 與 WebArena 達成 74.6% 推論成本降幅。" publishDate: "2026-05-03T15:00:00+08:00" updatedDate: "2026-05-03T15:00:00+08:00" tags: ["Computer-use Agents", "Yale NLP", "Inference Optimization", "GUI Automation"] draft: false ---

---
title: "【技術解析】讓電腦使用 Agent 不再昂貴：Yale NLP 的台階層級效率優化框架"
description: "Yale 與 UNC 研究者提出 event-driven 級聯架構，讓 AI Agent 在簡單步驟用小模型、風險時刻才叫用大模型，在 OSWorld 與 WebArena 達成 74.6% 推論成本降幅。"
publishDate: "2026-05-03T15:00:00+08:00"
updatedDate: "2026-05-03T15:00:00+08:00"
tags: ["Computer-use …

What You'll Learn How to build intelligent customer service agents that handle inquiries, resolve issues, and create exceptional customer experiences while knowing exactly when to bring in human support.

# 6.6 Customer Service Agents: Your 24/7 Customer Delight Team

## What You'll Learn
How to build intelligent customer service agents that handle inquiries, resolve issues, and create exceptional customer experiences while knowing exactly when to bring in human support.

## The H…

The "RENAPER workaround" pattern: orchestrate identity verification via providers that ARE accessible to indie devs (WhatsApp OTP, email magic-link, Auth0, MercadoPago Identity), get back a signed attestation with a trust level. The agent then decides if the trust level is enough for the action requested.

# @ar-agents/identity-attest: agent guide

The "RENAPER workaround" pattern: orchestrate identity verification via providers that ARE accessible to indie devs (WhatsApp OTP, email magic-link, Auth0, MercadoPago Identity), get back a signed attestation with a trust level. The agen…

See root `AGENTS.md` for monorepo-wide context.

# @commercetools-frontend/actions-global

See root `AGENTS.md` for monorepo-wide context.

## Purpose

Redux thunk action creators and React hooks for dispatching notifications and handling API errors in Merchant Center Custom Applications.

## Key Context

- This is the primary …

OS distribution assets for elizaOS: the elizaOS Live Linux USB distro (`linux/`) and the elizaOS AOSP Android fork (`android/`). This is a private, non-publishable package — it ships no JS runtime or exports consumed by other packages in the monorepo.

# @elizaos/distro-os

OS distribution assets for elizaOS: the elizaOS Live Linux USB distro (`linux/`) and the elizaOS AOSP Android fork (`android/`). This is a private, non-publishable package — it ships no JS runtime or exports consumed by other packages in the monorepo.

## Pu…

iMessage connector for Eliza agents on macOS — inbound polling via chat.db and outbound sending via AppleScript or a CLI tool.

# @elizaos/plugin-imessage

iMessage connector for Eliza agents on macOS — inbound polling via chat.db and outbound sending via AppleScript or a CLI tool.

## Purpose / role

Adds iMessage send/receive capability to an Eliza agent running on macOS. The plugin registers `IMessageS…

Pure TypeScript grid engine. Zero DOM dependencies. Provides the pipeline, sorting, filtering, grouping, pagination, cell editing, validation, export/import, infinite scroll, selection, tree-view, and i18n systems.

# @ornery/ui-grid-core — Agent Instructions

Pure TypeScript grid engine. Zero DOM dependencies. Provides the pipeline, sorting, filtering, grouping, pagination, cell editing, validation, export/import, infinite scroll, selection, tree-view, and i18n systems.

## Build & Test

``…

Summary

# A11y Audit per-Screen Checklist

## Summary

**One-sentence:** Per-screen accessibility audit checklist running 9 critical WCAG-AA checks (semantic landmarks, focus order, contrast, alt text, ARIA correctness, keyboard nav, error messaging, motion-respect, screen-reader labels)…

Status

# ADR-0014: GitHub Copilot Custom Agents with Three-Tier Architecture

## Status

Accepted

Date: 2025-01-06

## Context

As the statuspro-openapi-client project grew to include both a Python client library and
an MCP server, we needed a way to provide specialized AI assistance f…

Status: Accepted Date: 2026-05-18 Implements: D-24

# ADR-15: Restore doc-intake and eval-review agents

Status: Accepted
Date: 2026-05-18
Implements: D-24

## Context

ADR-07 (see `decisions/ADR-07-agent-trim.md`) dropped ten GSD agents during the v0.1.0 rationalization. Three of those agents (`gsd-doc-classifier`, `gsd-doc-synth…

**Generated:** 2026-02-02 **Commit:** f039639 **Branch:** main

# AFTER15 - PROJECT KNOWLEDGE BASE

**Generated:** 2026-02-02
**Commit:** f039639
**Branch:** main

## OVERVIEW

Rust CLI tool calculating overtime hours from Claude Code session logs. Parses JSONL session data, applies shift-based overtime rules (21-day rotating schedule), gener…

Build, Lint, and Test Commands

# AGENT DEVELOPMENT GUIDELINES

## Build, Lint, and Test Commands

### Running Tests
- Run all tests: `npm test`
- Run a single test file: `node --test test/security.test.js`
- Run specific test function: `node --test --test-name-pattern="securePath blocks paths outside CWD" test…

Overview - The main project code lives under `src/survaize`. - Unit tests are in `tests/` and evaluation scripts/data in `evals/`. - Development utilities such as lint scripts are under `devtools/`. - Documentation is primarily in `README.md`, `development.md`, and `installation.md`.

# AGENT Instructions

## Overview
- The main project code lives under `src/survaize`.
- Unit tests are in `tests/` and evaluation scripts/data in `evals/`.
- Development utilities such as lint scripts are under `devtools/`.
- Documentation is primarily in `README.md`, `developmen…

Tech Stack

# AGENT.md

## Tech Stack

- **TypeScript** (target: ES2022/ESNext)
- **pi-coding-agent**: Extension API for the pi harness.
- **Ollama REST API**: Communication with Ollama via HTTP requests (`fetch`).

## Entry Point: `src/index.ts`

- Exports a default function receiving `pi: …

<!-- BEGIN:nextjs-agent-rules --> # AGENT.md — Codex Instructions for SGE FILKOM UB Showcase

<!-- BEGIN:nextjs-agent-rules -->
# AGENT.md — Codex Instructions for SGE FILKOM UB Showcase

This file contains the working instructions for Codex or any AI coding agent that will build the **Website BEM FILKOM UB Showcase** interactive presentation website.

The project is a we…

Если ты выступаешь ревьювером (запущен через codex-review workflow): - Давай конкретный actionable фидбек - Можешь смотреть код/diff самостоятельно - Не запускай скрипты из skills/codex-review/ — ты ревьюер - Не заглядывай в .codex-review/archive/ — там артефакты прошлых сессий - После ревью запиши вердикт по пути, указанному в промпте ревью (одно слово: APPROVED или CHANGES_REQUESTED)

## Review Protocol

Если ты выступаешь ревьювером (запущен через codex-review workflow):
- Давай конкретный actionable фидбек
- Можешь смотреть код/diff самостоятельно
- Не запускай скрипты из skills/codex-review/ — ты ревьюер
- Не заглядывай в .codex-review/archive/ — там артефа…

If the user says to stop, says they will look, or otherwise indicates they want to pause — make NO further changes (no edits, no tool calls, no code) until explicitly told to continue.

## Interaction Rules

If the user says to stop, says they will look, or otherwise indicates they want to pause — make NO further changes (no edits, no tool calls, no code) until explicitly told to continue.

## Context

Python project for semantic ETL and reporting.

Provides a S…

- Use Conventional Commits for commit messages (e.g., `feat:`, `fix:`, `docs:`). - Keep changes minimal and scoped to the request. - Use Bun commands by default (`bun install`, `bun run ...`, `bunx ...`). - Keep docs/tooling phrasing implementation-agnostic where possible. - Don’t add npm-specific lockfile or workflow changes unless explicitly requested. - Validate changes with Playwright MCP where available.

# AGENTS

- Use Conventional Commits for commit messages (e.g., `feat:`, `fix:`, `docs:`).
- Keep changes minimal and scoped to the request.
- Use Bun commands by default (`bun install`, `bun run ...`, `bunx ...`).
- Keep docs/tooling phrasing implementation-agnostic where possib…

Package Manager

# AGENTS

## Package Manager

- Use Bun as the default package manager and script runner in this repository.
- Prefer bun run <script> over npm, pnpm, or yarn commands unless explicitly requested.

## Common Commands

- Build app: bun run build
- Build docs: bun run docs:build
- …

Quick commands - Activate env: `source distillation/bin/activate` - Verify env: `python scripts/check_env.py` - Run training (example): `python3 main.py --output_dir ./output --vllm_server_base_url http://localhost:8000 ...`

## Working agreements (for coding agents)

### Quick commands
- Activate env: `source distillation/bin/activate`
- Verify env: `python scripts/check_env.py`
- Run training (example): `python3 main.py --output_dir ./output --vllm_server_base_url http://localhost:8000 ...`

### CLI…

- When configuring formatters and linters (e.g. for neovim), favour calling them with 1. `uv` 2. `mise` 3. `bunx`, in this order. This applies to `pre-commit` hooks as well, if there is no pre-made hook available. - Use `prek` for pre-commit hooks instead of `pre-commit`. - Always use `uv` to run Python commands.

- When configuring formatters and linters (e.g. for neovim), favour calling them with 1. `uv` 2. `mise` 3. `bunx`, in this order. This applies to `pre-commit` hooks as well, if there is no pre-made hook available.
- Use `prek` for pre-commit hooks instead of `pre-commit`.
- Alway…

UI IMPLEMENTATION PROMPT You are building the frontend for a web-based classroom allocation and rescheduling system using Next.js (App Router), Tailwind CSS, shadcn/ui, Framer Motion, and Inter font. Follow these instructions strictly.

UI IMPLEMENTATION PROMPT
You are building the frontend for a web-based classroom allocation and rescheduling system using Next.js (App Router), Tailwind CSS, shadcn/ui, Framer Motion, and Inter font. Follow these instructions strictly.

1. Design System
Colors
Primary: Blue (shad…

When creating documentation, use the following rules: 1. Be succint. Do not be overly verbose. 2. Show only one example of each piece of functionality. 3. Provide a reason as to why you would want to use this functionality. 4. Avoid using emojis. 5. Write in a professional manner as if this were for enterprise users. 6. Create a brief, but useful, summary at the end. 7. Use named parameters in function calls if there is more than one parameter.

When creating documentation, use the following rules:
1. Be succint. Do not be overly verbose.
2. Show only one example of each piece of functionality.
3. Provide a reason as to why you would want to use this functionality.
4. Avoid using emojis.
5. Write in a professional manner…

--- name: pi-jobs description: Agent run telemetry and cost tracking for pi — auto-records every agent invocation with token usage, cost, duration, and tool call stats ---

---
name: pi-jobs
description: Agent run telemetry and cost tracking for pi — auto-records every agent invocation with token usage, cost, duration, and tool call stats
---

## Overview

Self-contained pi extension that auto-tracks all agent runs via lifecycle hooks. Stores data i…

--- name: pi-jobs description: Agent run telemetry and cost tracking for pi — auto-records every agent invocation with token usage, cost, duration, and tool call stats ---

---
name: pi-jobs
description: Agent run telemetry and cost tracking for pi — auto-records every agent invocation with token usage, cost, duration, and tool call stats
---

## Overview

Self-contained pi extension that auto-tracks all agent runs via lifecycle hooks. Stores data i…

Данный репозиторий нужен для автоматизации тестирования DHCP сервера в AdGuardHome Для разработки используется Typescript Внутри Docker контейнера поднимается QEMU который является гостем Нельзя использовать привелигированные контейнеры

Данный репозиторий нужен для автоматизации тестирования DHCP сервера в AdGuardHome
Для разработки используется Typescript
Внутри Docker контейнера поднимается QEMU который является гостем
Нельзя использовать привелигированные контейнеры

Project overview

# AGENTS

## Project overview

This repository contains the **answer**, **fetch-url**, **web-search**, **status**, **plan-md**, **task-subagents**, **review**, and **mention-skills** pi extensions.

## Key files

- `package.json`: npm workspace manifest for local linking of `shar…

- Remove all compiler warnings - Use info! log for non-frequent events and use warn! for errors - Never copy-and-paste code. Eliminate duplicated code - Run `cargo fmt` - Never use `allow(dead_code)`

## Style

- Remove all compiler warnings
- Use info! log for non-frequent events and use warn! for errors
- Never copy-and-paste code. Eliminate duplicated code
- Run `cargo fmt`
- Never use `allow(dead_code)`

## Testing

- Add appropriate tests
- Get `cargo test` pass
- Never a…

Never run builds or git commits

Never run builds or git commits

You are able to use the Svelte MCP server, where you have access to comprehensive Svelte 5 and SvelteKit documentation. Here's how to use the available tools effectively:

## Available MCP Tools:

### 1. list-sections

Use this FIRST to discover a…

<!-- SPECKIT START --> For additional context about technologies to be used, project structure, shell commands, and other important information, read: specs/007-pdf-textbox-fit/plan.md <!-- SPECKIT END -->

<!-- SPECKIT START -->
For additional context about technologies to be used, project structure,
shell commands, and other important information, read:
specs/007-pdf-textbox-fit/plan.md
<!-- SPECKIT END -->

The agent's role in this project is advisory only. The maintainer writes the code and configs himself; the agent's tasks are limited to:

## Working approach

The agent's role in this project is advisory only. The maintainer writes the code and configs himself; the agent's tasks are limited to:

- Hints, advice, and explanations of how and why things are done
- Reminders about missed steps, inconsistencies, and pot…

The agent's role in this project is advisory only. The maintainer writes the code and configs himself; the agent's tasks are limited to:

## Working approach

The agent's role in this project is advisory only. The maintainer writes the code and configs himself; the agent's tasks are limited to:

- Hints, advice, and explanations of how and why things are done
- Reminders about missed steps, inconsistencies, and pot…

High-signal facts for AI agents working in this repo. When in doubt, omit.

# AGENTS

High-signal facts for AI agents working in this repo. When in doubt, omit.

## Stack & setup

- **Express 5** (not 4 — `req.query.date` is `undefined` not `""` when param missing).
- **EJS v5** views in `src/views/` (set in `server.js`).
- **SQLite via sql.js** (WASM, i…