메뉴
How to approach Datadog investigations effectively — defer to the Datadog MCP's own skill guides for query syntax. Use when starting any Datadog logs/traces/metrics/RUM investigation.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Augusto's git commit, branch, and PR conventions. Use when writing commit messages, drafting PRs, or creating branches, to match a consistent style across repos.
How to add or modify a module (aspect) in this Den/flake-parts NixOS+darwin config. Use when creating a new program/service module, declaring flake inputs, or wiring an aspect into a host or user.
How to add or configure a Neovim plugin in this repo — flake input, build strategy, and lze lazy-loading lua config. Use when adding an nvim plugin, changing its build, or writing its plugin config.
How to add and manage secrets in this repo with sops-nix (age + GPG/YubiKey). Use when adding an API key/token/secret, editing the encrypted env.yaml, or exposing a secret to a program.
| name | datadog-queries |
| description | How to approach Datadog investigations effectively — defer to the Datadog MCP's own skill guides for query syntax. Use when starting any Datadog logs/traces/metrics/RUM investigation. |
The Datadog MCP ships its own extensive, always-current skill catalog
(list_datadog_skills returns ~90 domain guides with the real query syntax,
attributes, and footguns). Those are the source of truth for syntax — do
not hand-roll query DSL from memory. This skill only captures the workflow and
this org's conventions.
load_datadog_skill for the obvious
domain (datadog/logs, datadog/traces, datadog/metrics, datadog/rum,
datadog/incidents-and-alerting) and list_datadog_skills with fuzzy topic
keywords. Load matches + their related skills. Add datadog/visualizations
when charting.get_change_stories (what changed).aggregate_* to find the shape
(top services, error rate, p95 by endpoint) before pulling raw
events/spans/logs.env:prod, region tags)mcp.datadoghq.eu.Lead with the likely root cause + confidence, then evidence (queries + numbers)