「guard」の検索結果

Cite-or-Block 架构的基石原子 skill。把报告里的 citation 锚点 (例如 [guideline:CSCO-2024-NSCLC:§5.5.2] / [pmid:12345678:abstract]) 解析为源文件片段,并核对事实声明里的关键词是否在引用源原文里出现。 Foundation atomic skill for the Cite-or-Block architecture. Resolves citation anchors (e.g. [guideline:CSCO-2024-NSCLC:§5.5.2] / [pmid:12345678:abstract]) to raw source text fragments, and verifies whether claim keywords actually appear in the cited source. 使用场景 / Use when: 1. 报告生成后要做 fact-check (resolve_citation + verify_claim_against_source) 2. content-verification-layer 要扫全报告锚点 (parse_citations_in_text) 3. drug-citation-verifier / 任何 cross-check skill 需要"锚点 → 源原文" P0 守护:本 skill **绝不维护任何已知字典**。源里没有 = verified=False, 无例外、无 fallback、无"已知答案兜底"。 P0 guard: this skill **never maintains any built-in dictionary**. Not in source = verified=False. No fallback, no built-in answers.

日本語の概要は準備中です。原文の説明を表示しています。

Integrate authentication and authorization flows with provider-specific setup and RBAC safeguards.

日本語の概要は準備中です。原文の説明を表示しています。

GStack careful mode for risky work that needs a conservative pass, hazards, safeguards, and explicit confirmation points before execution.

日本語の概要は準備中です。原文の説明を表示しています。

Copy one Cloudflare D1 database into another of the same app's environments — refresh STAGING from prod, seed a PREVIEW from staging, or (guarded) push staging up to prod. Full mirror via scripts/db-clone.mjs. Use when asked to "clone a db", "refresh staging from prod", "copy prod data to staging", "seed a preview db", or to reconstitute a wiped staging DB. NOT for per-collection seeding (use the app's seedData/seed endpoints) or for schema migrations (use db-migrations).

日本語の概要は準備中です。原文の説明を表示しています。

The router for how a change legally happens in this repo — which sign-off gate fires for which kind of change, what counts as approval, how the packages-edit guard works, and how work is allowed to reach main. Use when starting ANY change and unsure which rules apply, when an Edit to packages/* gets BLOCKED, when a pipeline issue sits on status:blocked and you need to know what unblocks it, or when asked "do I need sign-off for this", "how do I get this merged", "can I deploy this", "who approves X", "why won't my approval label/reaction do anything".

日本語の概要は準備中です。原文の説明を表示しています。

运用《智能体设计模式》(Antonio Gulli) 的 21 个模式,指导 AI Agent 系统的架构设计与选型。适用于用户在设计 Agent 工作流(选择链式/路由/并行/多Agent编排)、提升 Agent 输出质量与可靠性(反思/异常处理/Guardrails)、做 RAG 技术选型(混合检索/GraphRAG/Agentic RAG)、优化 Agent 成本与延迟(动态模型切换/回退)时使用。

日本語の概要は準備中です。原文の説明を表示しています。

Classify + announce the target Convex deployment before any deployment-affecting command; fresh explicit consent for prod actions; session read-only mode.

日本語の概要は準備中です。原文の説明を表示しています。

Classify + announce the target Convex deployment before any deployment-affecting command; fresh explicit consent for prod actions; session read-only mode.

日本語の概要は準備中です。原文の説明を表示しています。

Design spec with 98 rules for building CLI tools that AI agents can safely use. Covers structured JSON output, error handling, input contracts, safety guardrails, exit codes, and agent self-description.

日本語の概要は準備中です。原文の説明を表示しています。

Autonomous DevSecOps & FinOps Guardrails. Orchestrates Gemini 3 Flash to audit Linux Kernel patches, Terraform cost drifts, and K8s compliance.

日本語の概要は準備中です。原文の説明を表示しています。

Detect and defend against prompt injection hidden in documents, web pages, and images consumed by an agent.

日本語の概要は準備中です。原文の説明を表示しています。

Triage npm packages for install-script malware, exfiltration, and worming behavior.

日本語の概要は準備中です。原文の説明を表示しています。

Deploy cloud-native deception across AWS, Azure, and GCP using decoy (honey) resources whose only purpose is to generate a high-fidelity alert the instant an attacker touches them: canary IAM access keys, permission-less decoy users/roles/service principals, honey object-storage buckets, and decoy secrets in Secrets Manager / Key Vault / Secret Manager. Wires detection through CloudTrail + EventBridge, Azure Sentinel honeytoken watchlists + Defender, and GCP Cloud Audit Logs, so any use of a decoy is routed to the SOC with near-zero false positives. Use when protecting cloud accounts and data stores, when an org has only on-prem honeypots and needs cloud coverage, when seeding fake AWS keys to catch credential theft and code-leak exposure, or when detecting cloud reconnaissance and lateral movement. Keywords: cloud deception, canary token AWS, honey S3 bucket, decoy IAM credentials, CloudTrail alert, GuardDuty, Sentinel honeytoken, decoy secret, honey service account, cloud honeypot, breach detection.

日本語の概要は準備中です。原文の説明を表示しています。

Hardware Security Modules (HSMs) are tamper-resistant physical devices that safeguard cryptographic keys and perform cryptographic operations in a hardened environment. Keys stored in an HSM never lea

日本語の概要は準備中です。原文の説明を表示しています。

Hardcoded-secret detection — AWS keys, GitHub PATs, Slack tokens, Stripe keys, generic high-entropy strings. Pre-commit (`--staged-only`), explicit paths, or full repo. Use when guarding `git commit`, auditing a repo for leaked credentials, validating no secrets entered the diff before push, or producing a rotation list for an exposed-secret incident.

日本語の概要は準備中です。原文の説明を表示しています。

内部由来の知見（ふりかえりの反映・実プロジェクトの検証記録など）を公開リポジトリに出す前に、組織固有の固有名詞（会社名・内部リポ名・顧客名・人物名・プロジェクト名・ID・パス等）の混入を独立点検して防ぐ。公開 git 履歴は遡れて消せないため、コミット / push の前に必ず通す。機密・認証情報の検査は扱わない。

대화형 System-wide Refactoring — 변경 파일에서 Extract Method·Domain Logic 이동·Pull Up 등 클래스 경계를 넘는 구조 변경 후보를 찾아 승인 후 기법별 커밋으로 적용. "구조 리팩토링", "도메인 로직 이동", "클래스 분리", "system-wide", "/system-wide-refactoring" 요청 시 사용. 단, 메서드 내부 정돈(guard clause·extract variable 등)은 /tdd-tidy가 적합. /system-wide-refactoring [commit-ref]로 호출.

日本語の概要は準備中です。原文の説明を表示しています。

Review+fix protocol with optional pre-core git diff review and safety guardrails (unsoundness, invariants, footguns, incidental complexity). Use when prompts say "/tw-fix this PR", "fix current branch", "fix this diff", "repair CI red", or "apply a minimal patch", and when crash/corruption/invariant-break issues need correction with a validation signal. Stop only after self-review exhausts actionable changes, each full cycle reaches post-self-review rerun plus terminal diff-review closure, and two consecutive zero-edit full cycles are clean.

日本語の概要は準備中です。原文の説明を表示しています。

Audit code, docs, tests, and registries against the Constitution, ADRs, and decision map without making broad repairs. Use when checking whether the repository is drifting away from accepted architecture.

日本語の概要は準備中です。原文の説明を表示しています。

Inspect a reference (web URL, image, PDF, screenshot) and propose a candidate Style for the workspace. Use when the user says "save this style", "guarda este estilo", or shares a reference and wants to reapply it later.

日本語の概要は準備中です。原文の説明を表示しています。

Repair confirmed architecture drift in small batches while preserving public canon and avoiding broad opportunistic refactors.

日本語の概要は準備中です。原文の説明を表示しています。

Load when working on Tier 0 Sovereign (payments, money, orders), Tier 1 Security (auth, tenancy, migrations), or Tier 2 Privacy (PII, erasure) tasks to enforce advanced model tier, Invariant-Breaker adversarial tests, and zero-PII telemetry validation.

日本語の概要は準備中です。原文の説明を表示しています。

Route protection with beforeLoad, redirect()/throw redirect(), isRedirect helper, authenticated layout routes (_authenticated), non-redirect auth (inline login), RBAC with roles and permissions, auth provider integration (Auth0, Clerk, Supabase), router context for auth state.

日本語の概要は準備中です。原文の説明を表示しています。

Abbruchbedingung für autonome Loops: So weit wie möglich selbständig weiterarbeiten UND einen Loop erst beenden, wenn belegt ist, dass keine autonom ausführbare Aufgabe mehr vorliegt. Der bloße Eindruck "nichts mehr zu tun" reicht NICHT — er löst eine vierstufige Prüfkette aus (think/decide, _CONTROL/_DECISIONS, Gardener/USMC, decision-avatar/BYUM), die zuerst neue Aufgaben GEWINNEN muss, bevor der Loop enden darf. Jeder Kettenschritt meldet found/empty/unavailable statt eines binären Ergebnisses — "exhausted" gilt nur, wenn ALLE Quellen tatsächlich befragbar waren; ist mindestens eine Quelle unavailable, meldet der Skill "blind" statt "exhausted" und behauptet nichts Ungeprüftes. Nutze diesen Skill bei /work-autonomous, /waafap, "arbeite autonom weiter bis nichts mehr zu tun ist", als Goal-Bedingung innerhalb eines /loop, oder wenn geprüft werden soll, ob wirklich keine autonome Arbeit mehr übrig ist, bevor ein Loop/eine Session beendet wird.

日本語の概要は準備中です。原文の説明を表示しています。