测试设计、漏洞检测、质量保证（1.3万 个）
Autonomous model validation and benchmarking. Investigates any ML model (LLM, image gen, TTS, time series, etc.), runs it on GPU cloud, evaluates quality and performance, and generates HTML reports. Use when user asks to verify, benchmark, evaluate, or test a model. Triggers on "verify model", "benchmark", "evaluate model", "test model", "run benchmark", "model evaluation", "モデルを検証", "ベンチマーク", "モデルを試して".
中文简介正在准备中。当前显示原文说明。
Create a construction plan for a multi-step engineering task that requires multiple PRs or a phased approach. Produces a self-contained plan file with dependency graph, review gates, and step-by-step execution instructions for coding agents. TRIGGER when: user requests a plan, blueprint, or roadmap for a complex task, or describes work that clearly needs multiple PRs. DO NOT TRIGGER when: task is completable in a single PR or fewer than 3 tool calls, user asks to modify/update an existing plan, user says "just do it", or user asks for code review/audit.
中文简介正在准备中。当前显示原文说明。
Show or change Peephole's enforcement mode (guided (default), audit, strict, hardened, off). Use when the user says /peephole, wants to see the current security mode, or asks to make enforcement stricter or looser.
中文简介正在准备中。当前显示原文说明。
Review the current changes for the CWE classes Peephole enforces, as a checklist including the absence-of-guard classes regexes cannot catch. Use when the user says /sec-review or asks for a security review of recent code before committing.
中文简介正在准备中。当前显示原文说明。
Explain and list Peephole sec-debt markers — deliberate, human-authorized security trade-offs recorded in code. Use when the user says /sec-debt, wants to knowingly accept a security finding, or asks what deferred security items exist.
中文简介正在准备中。当前显示原文说明。
Print Peephole's security audit report for this project — current mode, binary integrity, the CWE-mapped decision tally, sec-debt items needing approval, and tamper-evident log-chain status. Use when the user says /sec-audit or asks what Peephole has blocked or flagged.
中文简介正在准备中。当前显示原文说明。
Personal App-in-Skill customer-support desk (help desk) that triages incoming support tickets from email, WhatsApp, web chat, contact forms, and WeChat, drafts KB-grounded replies, and proposes actions, then holds every send behind a human approval queue with a pre-send quality gate (support-qa) that outputs SHIP/FIX/BLOCK, plus SLA and CSAT tracking. Use when the user invokes $kelly-support or /kelly-support, mentions 客服 / 工单 / 支持台, a support desk, help desk, customer service, support tickets, ticket triage, a knowledge base or canned macros, SLA / first-response time, CSAT, escalations, or refund requests, wants incoming tickets triaged into a needs_review → approved → done pipeline, agent-drafted replies grounded in a knowledge base and reviewed before sending, refunds and commitments gated behind human approval, or breaching-SLA tickets surfaced. This is the OPERATOR desk (agent + human), distinct from a visitor-facing chat bubble.
中文简介正在准备中。当前显示原文说明。
Publish agent skills AND MCP servers to every marketplace, plugin directory, and registry — security-scan for private data, validate with `gh skill`, cut a release, wire the Claude /plugin and Codex marketplaces, publish an MCP server to the official MCP Registry (mcp-publisher + server.json), and prep the curated / monetized stores (Buda Marketplace, Agensi, OpenClaw ClawHub) and review-gated directories (ChatGPT App Directory, Anthropic Connectors + Desktop Extensions/.mcpb). Use when the user wants to publish, release, list, 上架, or distribute agent skills, plugins, or MCP servers to skills.sh / SkillsMP / Claude Code / Codex / the MCP Registry.
中文简介正在准备中。当前显示原文说明。
Design and create Busabase-backed App-in-Skill packages with a canonical app project, Busabase-native setup and product onboarding, review/execution workflows, a standardized responsive UI, and AirApp-first delivery. Use when a user wants a Busabase research desk, review queue, planner, action console, operating dashboard, control panel, collaboration workspace, onboarding/readiness flow, or an existing Kelly App-based skill updated. Every generated skill contains a complete app/ project, deploys that source to Busabase AirApp by default, runs pnpm dev only when local preview is explicitly requested, follows the Kelly desktop and phone UI contract, delegates AirApp runtime, SDK, security, scaffolding, and deployment constraints to busabase-app-creator, and delegates repository-level conformance testing to kelly-app-skill-creator-tests.
中文简介正在准备中。当前显示原文说明。
Personal finance anomaly review desk (Busabase App-in-Skill) that imports the three business tables — orders, invoices, payments/receipts (回款) — from CSV/JSON exports, flags anomalies with deterministic rules, and runs a human review queue with auditable decisions. Use when the user invokes $kelly-audit or /kelly-audit, mentions finance audit, order invoice payment reconciliation, 对账, 应收, 回款, 发票异常, receivables aging, overdue receivables, duplicate payments, missing invoices, amount mismatches, or anomaly review.
中文简介正在准备中。当前显示原文说明。
Busabase App-in-Skill review queue that scores candidate SME financing deals (revenue-based/RBF-style credit) with a deterministic, fully auditable rule-based rubric — never an LLM or API call. Use when the user invokes $kelly-deal-scorer or /kelly-deal-scorer, wants to review a deal-underwriting queue, score financing candidates, compute a composite score breakdown, see a suggested revenue-share rate range, or record approve/send-back/reject decisions for a private-credit or RBF-style lending pipeline.
中文简介正在准备中。当前显示原文说明。
Build, maintain, and run conformance tests for canonical App-in-Skill projects created by kelly-app-skill-creator. Use when a Kelly app skill needs contract checks, local server smoke tests, responsive browser acceptance, temporary open-source Busabase integration, environment-gated Busabase Cloud OAuth verification, lazy provisioning checks, persistence tests, AirApp parity tests, or CI coverage. Keep app creation in kelly-app-skill-creator and own only the reusable test harness, per-skill fixtures, execution gates, diagnostics, and truthful pass/fail/skip reporting.
中文简介正在准备中。当前显示原文说明。
Digital-human implementation and demo skill for choosing, prototyping, and QA-ing low-cost 2D photoreal digital humans and high-control 3D custom digital humans. Use when the user wants a digital-human avatar, AI host, customer-service presenter, multimodal voice-to-lip-sync video stream, vendor comparison for services such as Silicon Intelligence / Tencent Zhiying / ZEGO-like real-time providers, or a UE/Unity 3D digital-human architecture and launch-ready demo.
中文简介正在准备中。当前显示原文说明。
Review board (Busabase App-in-Skill) that runs a fixed suite of mock test cases against a baseline vs candidate agent version and surfaces rubric-scored regressions before a release. Use when the user invokes $kelly-agent-eval or /kelly-agent-eval, wants to review agent-version regressions, compare baseline vs candidate quality, triage a release, or record a release approve/block decision. Deterministic mock rubric scores only — not a real LLM-judge call, and it never deploys anything.
中文简介正在准备中。当前显示原文说明。
Outbound email-marketing desk (Busabase App-in-Skill) for building segments, drafting campaigns, newsletters, and sequences, running pre-send deliverability and subject-line QA, and approving every send before it is scheduled. Structured around the SEND discipline — Setup, Engage, Nurture, Deliver — with an email-quality-auditor gate (EQS score + SHIP/FIX/BLOCK verdict). Use when the user invokes $kelly-campaigns or /kelly-campaigns, or mentions email marketing, campaigns, newsletters, drip/welcome/win-back sequences, broadcasts, segments, subject-line A/B tests, deliverability (SPF/DKIM/DMARC/spam score/inbox placement), or wants to review and approve marketing email before it is sent. This is OUTBOUND marketing to a subscriber list, distinct from kelly-email inbox triage (incoming mail). 出站邮件营销：策划分群、起草营销活动 / 新闻邮件 / 邮件序列，发送前做可送达性与主题行质检，人工审批后再排期发送。
中文简介正在准备中。当前显示原文说明。
Apply repeatable dialogue MA, Studio-based or automatic portrait reframing, overlay-safe rendering, and loudness/sync QA to an existing interview or talking-head rough cut. Use when the user asks to make speech easier to hear, perform MA, enlarge or rebalance a person in frame, crop or punch in an interview, automatically propose framing, or prepare a finishing preview before final approval.
中文简介正在准备中。当前显示原文说明。
Select an evidence-backed platform delivery profile, preserve organic and ads scope boundaries, and produce safe-zone QA input with an explicit human preview gate when evidence is unknown or stale.
中文简介正在准备中。当前显示原文说明。
Use when reviewing or writing LLM, RAG, MCP, tool, or agent code for OWASP-aligned security issues; triggered by "owasp my code", "owasp this PR", AI security review, PR review, or changes to AI system code.
中文简介正在准备中。当前显示原文说明。
Audit a browser fingerprint for internal contradictions with the liarjs CLI - canvas, WebGL, WebGL2, WebGPU, audio, 220 fonts, WebRTC and timezone probes, scored against the TLS/HTTP/ASN view of the same request. Use when asked to run a browser fingerprint test, see what a fingerprint looks like, check canvas or WebGL fingerprint stability, compare a spoofed profile against a real browser, or find out whether a browser profile is self-consistent.
中文简介正在准备中。当前显示原文说明。
Check whether a Playwright, Puppeteer, Selenium or CDP-driven browser presents a coherent fingerprint, using liarjs as a library against a Page you already have - navigator.webdriver, HeadlessChrome tokens, worker versus main-thread identity, patched-API integrity, WebGL versus WebGPU GPU identity. Use when asked whether an automated browser looks like a normal one, when a headless setup or a stealth plugin's effect needs measuring rather than assuming, or when an assertion on fingerprint quality belongs in a test suite.
中文简介正在准备中。当前显示原文说明。
Add a new agent to this AgentOS. Runs guided discovery or takes a concrete idea, then generates agents/slug.py, registers it in app/main.py, adds its manifest entry (description + quick prompts), reloads the selected development runtime, and smoke-tests it for you. Covers the product agent too — ingest a product's website/docs into the product knowledge base and build a knowledge-only agent over it. Use whenever the user wants to add or create a new agent, or names a product (theirs or one they use) they want an agent for.
中文简介正在准备中。当前显示原文说明。
The method for getting a cheaper model to match an agent's big-model answers by fixing instructions, inputs and setup. Use when asked to move an agent to a smaller or cheaper model, prove one model beats another, or compare models on an agent's test cases.
中文简介正在准备中。当前显示原文说明。
Vite (フロントエンドビルドツール・開発サーバー, v8 / Rolldown) のリファレンス。 vite.config, defineConfig, server, build, optimizeDeps, HMR (import.meta.hot), import.meta.env, import.meta.glob, configureServer, transformIndexHtml, Environment API (DevEnvironment / ModuleRunner), SSR, library mode, JavaScript API, v7 から v8 移行。 テストは vitest スキル。
中文简介正在准备中。当前显示原文说明。
Electron (Chromium + Node.js 製クロスプラットフォームデスクトップアプリフレームワーク, v44) の リファレンス。main / renderer / preload / utilityProcess, ipcMain / ipcRenderer / contextBridge, BrowserWindow / BaseWindow / WebContentsView / webContents, Electron Forge, autoUpdater, safeStorage, fuses, ASAR, contextIsolation / sandbox, code signing, MAS, crashReporter, contentTracing, debugging, E2E testing, breaking changes。
中文简介正在准备中。当前显示原文说明。