Complex data initiatives require coordinated legal, security, privacy, policy, standards, and sector review. This skill organizes those perspectives into evidence-based assessments and remediation plans.

Cloudflare One deployments span identity, traffic, devices, private networks, and data controls. This skill structures design, configuration, troubleshooting, validation, and staged rollout decisions.

Firestore Rules can expose private data when access and validation logic are incomplete. This skill helps create structured rules with least-privilege access, schema checks, and security test scenarios.

Browser login automation can expose credentials when agents handle them directly. This skill provides a consent-based Bitwarden Agent Access workflow with domain checks, sanitized status, and explicit trust boundaries.

Challenge authors need to reduce shallow LLM pattern matching without making fair challenges frustrating. This skill reproduces the intended solve, selects a small resistance change, and verifies the result end to end.

Security reviews can miss trust boundaries and unsafe data flows. This skill supplies threat-modeling prompts, OWASP patterns, and practical review checklists.

Skill authors need reliable checks for package safety, triggering, behavior, and release readiness. Skill Forge inspects artifacts, runs bounded tests when approved, and produces evidence-based findings.

Security reviews require consistent scope, evidence, and remediation guidance across many tools. This skill provides practical workflows for authorized assessment, monitoring, hardening, and reporting.

Fraud teams need practical methods to identify abuse, test controls, and monitor drift. This skill guides risk analysis, rules, models, backtesting, and real-time system design.

雲端伺服器設定經常失敗，原因是 SSH、防火牆、DNS、Nginx 和 TLS 步驟順序錯誤。此技能會引導 Claude、Codex 或 Claude Code 完成經過驗證、符合發行版差異的主機代管與強化工作流程。

如果雲端存取設定得過於寬鬆，自架的 OpenClaw 可能會暴露強大的控制平面。此技能會引導採用隱私優先的 Linux 部署，涵蓋強化的主機設定、loopback 綁定、SSH 通道、Podman、權杖驗證、配對，以及沙箱預設值。

Manual image review is slow and inconsistent. This skill compresses image batches, submits them to an MCP moderation service, and summarizes pass, blocked, and failed results.

智能合約團隊在部署協議變更前需要可重複執行的測試。此技能提供實用的 Hardhat 與 Foundry 模式，涵蓋單元測試、分叉、模糊測試、gas 檢查、覆蓋率與驗證。

安全團隊需要一種可重複的方法，在設計交付前識別威脅。此技能會引導 Claude、Codex 和 Claude Code 依循 STRIDE 類別、風險評分與緩解措施文件化流程進行分析。

威脅模型通常停留在風險層面，缺乏明確的實作工作。此技能會將威脅、合規需求與業務脈絡轉換為可測試的安全需求。

CI/CD 管線經常透過硬編碼值、記錄檔或薄弱的存取控制暴露認證資訊。此技能可協助 Claude、Codex 和 Claude Code 設計更安全的機密儲存、擷取、輪替和掃描工作流程。

安全團隊需要能在不同工具與儲存庫之間重複使用的 SAST 設定。此技能會引導 Claude、Codex 和 Claude Code 完成設定、調校與 CI/CD 整合。

支付系統必須保護持卡人資料並縮小合規範圍。此技能可協助 Claude、Codex 和 Claude Code 以實務實作指引套用 PCI DSS 控制措施。

服務團隊需要在服務網格與 Kubernetes 叢集之間採用一致的 mTLS 模式。此技能提供實用範本、輪替指引，以及安全服務通訊的疑難排解步驟。

事件團隊需要在服務中斷發生前就具備清楚的應變步驟。此技能可協助 Claude、Codex 和 Claude Code 起草結構化 runbook，包含分流、緩解、升級與溝通章節。