Use this skill when you need to search Datadog logs, query metrics, tail logs in real-time, trace distributed requests, investigate errors, compare time periods, find log patterns, check service health, or export observability data.
Skills(SKILL.md)は、AIエージェント(Claude Code、Cursor、Codexなど)に特定の能力を追加するための設定ファイルです。
詳しく見る →Use this skill when you need to search Datadog logs, query metrics, tail logs in real-time, trace distributed requests, investigate errors, compare time periods, find log patterns, check service health, or export observability data.
Datalog bottom-up fixpoint iteration for recursive queries
date-timezone
Identify unused code, imports, variables, and functions for safe removal.
Use to manage pricing, packaging, and approval workflows for renewal
Score deal velocity based on email response times, meeting frequency, and stakeholder engagement. Predict which deals will close vs stall.
End-of-session debrief to capture learnings and keep documentation current.
decay-scoring
Decision modeling using decision tables, weighted scoring matrices, and decision trees. Structures complex decisions with clear criteria, alternatives evaluation, and outcome prediction.
Validate implementation approaches against Brief decisions
Capture architectural and implementation decisions with rationale to prevent re-debating. Creates and maintains DECISIONS.md.
A decision matrix is a structured tool for comparing multiple alternatives against weighted criteria to make transparent, defensible choices. It forces explicit trade-off analysis by scoring each opti
Create or update an audit of recent decisions for quality
Create or update a structured decision record
Facilitate data-driven technical decisions using embedded decision matrices and trade-off analysis. Use when relevant to the task.
Git-powered state awareness - track file changes, session context, decisions, and work history. Query what happened in previous sessions and during current session. Auto-activates for state queries and before duplicating work.
Strategic decision-making for CFN Loop progression with robust parsing. Use when evaluating validator consensus and determining PROCEED/ITERATE/ABORT outcomes.
Use when starting plan execution to decompose implementation plans into Claude Code Task tool tasks at H2 (Task) granularity - creates one task per plan task with citation-manager directives, keeping plan as source of truth
Use after writing-plans to decompose monolithic plan into individual task files and identify tasks that can run in parallel (up to 2 subagents simultaneously)
Investiga documentación oficial y valida soluciones en internet antes de implementar.
Deep learning-based variant calling with Google DeepVariant. Provides high accuracy for germline SNPs and indels from Illumina, PacBio, and ONT data. Use when calling variants with DeepVariant deep learning caller.
Creates and manages multi-step AI workflows. Use when defining, implementing, or improving DeepWork jobs.
<!-- PM-Skills | https://github.com/product-on-purpose/pm-skills | Apache 2.0 -->
Delete files from knowledge base with directory governance. Use when user wants to remove or delete content. Reads RULE.md to check deletion permissions and handles dependencies.
container-use環境(コンテナ、ファイル、メタデータ)を完全に削除する手順
Use when investigating inbox placement, reputation, and compliance signals
Delta Sharing configuration, monitoring, and recipient management for secure cross-organization data sharing.
Analyze project dependencies for security vulnerabilities, outdated
Analyze project dependencies for vulnerabilities, updates, and optimization opportunities. Use when auditing dependencies or managing package versions.
Automated dependency management with security scanning, update orchestration, and compatibility validation
Add new services to DIContainer, create protocol-based implementations, set up singletons or factories for Leavn app dependency injection
>
Manage Python dependencies using uv package manager. Use when installing, adding, removing, or updating packages. Always prefer uv over pip.
Map, track, and manage project dependencies across teams, systems, and organizations. Identify critical path items and prevent blocking issues through proactive dependency management.
Health checks, metrics, alerting ve rollback stratejileri.
deploy-service
Deployment workflow for company repos (pre/pro) using CI, ECR, and Helm.
Set up progressive canary deployments on GCP Cloud Run with traffic splitting, monitoring alerts, and automated rollback.
Production deployment configurations for elizaOS agents with Docker, monitoring, and scaling.
Production deployment principles and decision-making. Safe deployment workflows, rollback strategies, and verification. Teaches thinking, not scripts.
Azure Deployment Stacks GA 2025 features for unified resource management, deny settings, and lifecycle management
Deployment Strategy Agent. Feature Flag, Canary/Blue-Green 배포, 롤백 전략을 담당합니다. 안전한 배포와 점진적 롤아웃을 관리합니다.
**Purpose**: Systematically validate the application is ready for production deployment, catching common "works locally, fails on Render" issues BEFORE they happen.
deployment-verification
Inline orchestration workflow for dependency audit and updates. Provides step-by-step phases for dependency-auditor detection, priority-based updates with dependency-updater, and verification cycles.
多方案評估決策框架. Use for: (1) 面臨 3+ 技術方案時的結構化評估, (2) 架構決策時的系統化分析, (3) 防止衝動決策和技術債務累積
Deep knowledge of legendary designers and their enduring contributions. Learn from Saul Bass, Massimo Vignelli, Dieter Rams, Paula Scher, and others whose work defines excellence. Use when seeking inspiration, understanding design history, or applying proven approaches.
Auto-activates when synthesizing outputs from multiple parallel sub-agents
Selects the appropriate quasi-experimental method (DiD, ITS, SC) based on data structure and research questions. Use when the user is unsure which method to apply.
Detect empty catch blocks and exception swallowing anti-patterns. Finds catch blocks that silently ignore errors without logging or rethrowing. Works across all languages. Use when user asks about error handling issues.