Recently added agent tools.

Validate and fix folder structure for Output SDK workflows. Use to ensure migrated workflows follow the correct structure conventions.

Use this to create or modify architecture decision records (ADRs)

Half-Quadratic Quantization for LLMs without calibration data. Use when quantizing models to 4/3/2-bit precision without needing calibration datasets, for fast quantization workflows, or when deploying with vLLM or HuggingFace Transformers.

Takes an approved plan from docs/plans/ and drives it to an open PR. Implements it by milestones, commits per milestone in the repo's style, and opens the PR against the repo's base branch. Use whenever the user approves a plan, says "ship it", "go", "start", "implement it", "do it", or asks to open a PR for already-planned work — or in Spanish "dale", "arrancá", "implementalo", "hacelo".

> Mechanically audit whether a project's harness is load-bearing. Orchestrates the deterministic check-harness-strength engine; interprets results. Not a deep/AI review. ## When to Use - On a milestone gate, to confirm the harness still constrains rather than merely decorates the project - When adding the strength audit as a required CI check, so the seven STRENGTH patterns block the merge instead of being prose advice - When validating a harness-distribution (toolkit) repo before shipping templ

Analyze PyTorch profiler and vLLM trace files with Perfetto SQL evidence. Use when working with torch profiler traces, merged_trace files, vLLM profiling logs, Perfetto trace analysis, GPU kernel hotspots, synchronization waits, rank imbalance, prefill/decode bottlenecks, or optimization reports.

soneta-ui-style is a design system that provides a consistent visual style for web applications following the Soneta/enova365 ERP interface guidelines, including color palettes, typography, layouts, UI components, and theming for both light and dark modes.

A GLSL library for creating 2D/3D procedural shapes and raymarched scenes using signed distance functions and boolean operations.

A multi-agent collaborative video creation and management tool that structures workflows, quality control, and data feedback to produce high-quality short videos with coordinated roles for content, visuals, audio, and quality assurance.

|

Handle social media crises with strategic response and reputation management. Use when the user needs to manage a PR crisis, handle negative feedback storms, or protect brand reputation.

d3k assistant for debugging web apps

Review and triage semgrep security scan results to identify true positive vulnerabilities. Use when analyzing semgrep output, triaging security findings, reviewing static analysis results, or when the user has semgrep-results directories to review. Performs deep code analysis to distinguish real vulnerabilities from false positives with high confidence.

Develop native Android apps with Kotlin. Covers MVVM with Jetpack, Compose for modern UI, Retrofit for API calls, Room for local storage, and navigation architecture.

Follow project-wide development conventions including consistent project structure, clear documentation, version control best practices, environment configuration, and dependency management. Use this skill when organizing project files and directories, writing README files, creating commit messages, managing environment variables, or handling dependencies. When working with configuration files, documentation files, package managers, or version control. When setting up new features, organizing co

Google Ads platform expertise. Audit campaigns, keywords, audiences, bidding, and conversion tracking. Use when the user asks about Google Ads, PPC, search advertising, Performance Max, Shopping ads, display campaigns, YouTube ads, or paid search optimization.

Use this skill when implementing accessibility for a specific Gamut component, building a custom overlay or composite widget, or auditing component usage against WCAG — complements the always-loaded `accessibility.mdc` with Gamut component-specific patterns. Form wiring lives in `gamut-forms`; Modal/Dialog composition detail lives in `gamut-modal`.

Use when you need to resolve an in-progress git merge/rebase conflict.

An enterprise-grade code quality orchestration tool that automates code review, validates quality gates, enforces best practices, and integrates proactive analysis within CI/CD pipelines.

click-scaffolder is a tool for generating Python CLI applications using Click, providing scaffolding with best practices and project structure.

Automates and orchestrates terminal tmux windows for running, monitoring, and managing background and interactive CLI tasks.

1 远程重启计算机 ``` echo b > /proc/sysrq-trigger ``` 2 查看内存使用情况 ``` free -m //以MB为单位查看内存使用情况 ``` 3 修改计算机的可用内存 对于Redhat系(包括Fedora) 修改/etc/grub.conf，在kernel一行最后添加mem=1G就能够将可用内存修改为1GB ``` title Red Hat Enterprise Linux Server (2.6.27) root (hd0,1) kernel /vmlinuz-2.6.27 ro root=/dev/VolGroup00/LogVol00 rhgb quiet acpi=ht mem=1G initrd /initrd-2.6.27.img ``` 4 查看系统的IO统计信息 ``` iostat -k 3 //以KB打印结果，3秒显示一次 ``` 5 将一个进程放入后台，并使它不受前台的影响，特别是用于远程连接时 ``` ./test //运行程序 ^Z //ctrl+z使程序暂停 bg %(jobid) //将程序放入后台运行 disow

Fully autonomous QA testing agent for Modular RAG MCP Server. Reads test cases from QA_TEST_PLAN.md, executes ALL test types automatically without human intervention — CLI commands, Dashboard UI via Streamlit AppTest headless rendering, MCP protocol via subprocess JSON-RPC, provider switches, and data lifecycle checks. Diagnoses failures, applies fixes with up to 3 retry rounds, and records results in QA_TEST_PROGRESS.md. Use when user says 'run QA', 'QA test', 'QA 测试', '执行测试', '跑测试', 'test and

Fully autonomous QA testing agent for Modular RAG MCP Server. Reads test cases from QA_TEST_PLAN.md, executes ALL test types automatically without human intervention — CLI commands, Dashboard UI via Streamlit AppTest headless rendering, MCP protocol via subprocess JSON-RPC, provider switches, and data lifecycle checks. Diagnoses failures, applies fixes with up to 3 retry rounds, and records results in QA_TEST_PROGRESS.md. Use when user says 'run QA', 'QA test', 'QA 测试', '执行测试', '跑测试', 'test and

Fully autonomous QA testing agent for Modular RAG MCP Server. Reads test cases from QA_TEST_PLAN.md, executes ALL test types automatically without human intervention — CLI commands, Dashboard UI via Streamlit AppTest headless rendering, MCP protocol via subprocess JSON-RPC, provider switches, and data lifecycle checks. Diagnoses failures, applies fixes with up to 3 retry rounds, and records results in QA_TEST_PROGRESS.md. Use when user says 'run QA', 'QA test', 'QA 测试', '执行测试', '跑测试', 'test and

An autonomous quality assurance testing agent that executes and diagnoses all types of tests without human intervention, applying fixes and recording results automatically.

Fully autonomous QA testing agent for Modular RAG MCP Server. Reads test cases from QA_TEST_PLAN.md, executes ALL test types automatically without human intervention — CLI commands, Dashboard UI via Streamlit AppTest headless rendering, MCP protocol via subprocess JSON-RPC, provider switches, and data lifecycle checks. Diagnoses failures, applies fixes with up to 3 retry rounds, and records results in QA_TEST_PROGRESS.md. Use when user says 'run QA', 'QA test', 'QA 测试', '执行测试', '跑测试', 'test and

Elite code review expert specializing in modern AI-powered code

A/B Rollout Governance for product management execution. Use when work requires a/b rollout governance for product management execution with guardrails, traceable execution, and measurable outcomes.

Provides expertise for analyzing DWARF debug files and understanding the DWARF debug format/standard (v3-v5). Triggers when understanding DWARF information, interacting with DWARF files, answering DWARF-related questions, or working with code that parses DWARF data.

Clean editorial-style interfaces. Warm monochrome palette, typographic contrast, flat bento grids, muted pastels. No gradients, no heavy shadows.

Optimize Bazel builds for large-scale monorepos. Use when configuring Bazel, implementing remote execution, or optimizing build performance for enterprise codebases.

Official Android tool that signs and verifies APKs against platform signature-scheme requirements.

Use when working with context management context restore

Apply STRIDE methodology to systematically identify threats. Use when analyzing system security, conducting threat modeling sessions, or creating security documentation.

Use when working with error debugging multi agent review

中文漂移审计技能。用于项目或学习过程变乱、上下文漂移、任务分叉、多个方案冲突、命名不一致、Codex 可能顺手改多了时。触发语包括"漂移检查""感觉跑偏了""项目变乱了""检查是否失控""分叉太多""上下文漂移"。

Azure Monitor OpenTelemetry Exporter for Java. Export OpenTelemetry traces, metrics, and logs to Azure Monitor/Application Insights.

Track ML experiments with Trackio using Python logging, alerts, and CLI metric retrieval.

|

|

Make any repo AI-ready — analyzes your codebase and generates AGENTS.md, copilot-instructions.md, CI workflows, issue templates, and more. Mines your PR review patterns and creates files customized to your stack. USE THIS SKILL when the user asks to "make this repo ai-ready", "set up AI config", or "prepare this repo for AI contributions".

**Tier:** POWERFUL **Category:** Engineering **Domain:** Multi-Agent Systems / AI Orchestration --- ## Overview Design production-grade multi-agent workflows with clear pattern choice, handoff contracts, failure handling, and cost/context controls. ## Core Capabilities - Workflow pattern selection for multi-step agent systems - Skeleton config generation for fast workflow bootstrapping - Context and cost discipline across long-running flows - Error recovery and retry strategy scaffolding - Docum

|

>

Design and implement a complete ML pipeline for: $ARGUMENTS

## One-line decision Use this skill when you want to estimate the influence of individual training examples on model predictions for data debugging and selection. Avoid it when you cannot afford the compute for influence estimation or have too many training examples. ## Skill metadata - **Skill type**: influence-based-data-valuation - **Paper kind**: operational-method - **Actionability**: medium - **Evidence quality**: full_paper ## Goal Use influence functions to estimate the effect of individ

## Overview This skill defines the authoritative pedagogical standards and workspace architecture for the CS Wiki (`cs-first-principles`) project. ## Directory Structure & Folder Roles - `platform/`: Modern CS Wiki Web Application (React + Vite + TypeScript) containing interactive simulators, knowledge graph, complexity-calibrated question bank, and zoomable abstraction layers. - `cs-learning-platform.md`: Authoritative 20-Point Pedagogy Blueprint (Concept Maps, Interactive Diagrams, Animations,

Search for cheap flights to Canada from US cities, with comprehensive research methodology and booking strategies.

Builds reusable Convex components with isolated tables and app-facing APIs.