SkillFoxx

Pup CLI

Pup is a Datadog CLI for agents: metrics, logs, monitors and traces by command, JSON output, skills for Claude Code and Codex.

  • 1,043 ★ on the repository
  • Last commit
  • Apache-2.0
CLI

Extended access

Why this level

  • Through the token the agent reads logs and metrics and can delete monitors, dashboards and SLOs if token rights allow it.
All reasons and checks

datadog/pupAuthor:DataDog

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add cli/pup

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • Through the token the agent reads logs and metrics and can delete monitors, dashboards and SLOs if token rights allow it.

This is third-party code. Review the repository files before installing.

What it does

Pup is a Rust command-line program from Datadog that opens the observability platform API to agents. Commands print structured JSON or YAML and describe themselves: `pup monitors list --tags="team:api-platform"`, `pup logs search --query="status:error" --from="1h"`, `pup metrics query --query="avg:system.cpu.user{*}"`. It covers metrics, logs, events, RUM, APM, traces, monitors, dashboards, SLOs, Synthetics, downtimes, security and other areas. Login uses OAuth2 with PKCE (`pup auth login`) instead of long-lived keys, and API keys can be set instead. The `pup skills install` command installs skills and 48 domain agents for Claude Code, Cursor, Codex, opencode and pi, such as `dd-pup`, `dd-monitors`, `dd-logs` and `dd-apm`. The `pup setup` command walks through connecting Datadog to a project (APM, Docker, ECS, Error Tracking) and in `--headless --recommend-only` mode prints JSON recommendations without changing anything.

Who it is for. Engineers whose monitoring lives in Datadog and who want an agent to search errors in logs, metrics and traces.

Good fit when

  • After a release the error rate climbs and the agent must find log records for the last hour.
  • You need the list of team monitors and which of them are alerting now.
  • An investigation goes through APM traces: from a service to dependencies and bottlenecks.
  • A project must be connected to Datadog and `pup setup` suggests the first step.

Not a fit when

  • You have no Datadog account: commands return nothing without it.
  • You need profiling: pup lacks it, and Datadog provides a separate MCP server for that.
  • Agent access to deleting monitors and dashboards is unwanted: such commands exist.

Example request

Use `pup logs search` to find checkout service errors for the last hour and list the most frequent messages.

Limitations

A Datadog account and login through `pup auth login` are required, and the site is set by a variable, for example `datadoghq.eu`. Some API areas are marked as not implemented in the README table: profiling is meant to go through the Datadog MCP server. Commands to delete monitors, dashboards and SLOs are included, so the agent should get read-only rights. The install script is fetched from the network, and `pup setup` runs the Datadog AI setup tool through `npx` and needs Node.js 22 or newer. The license is Apache 2.0, the last commit dates from November 2026.

How to disable. Remove the `pup` binary from `~/.local/bin` or run `brew uninstall pup`, then `npm uninstall -g @datadog/pup` if you installed through npm. Delete agent skills from their directories.

FAQ

How do I install Pup?

With the script `curl -fsSL https://raw.githubusercontent.com/DataDog/pup/main/install.sh | sh`, through `brew install datadog-labs/pack/pup` or `npm install -g @datadog/pup`.

How do I log in to Datadog?

With `pup auth login`: browser OAuth2 with PKCE and automatic token refresh. The alternative is API keys.

How do I install agent skills?

Run `pup skills install claude` (or `codex`, `cursor`, `all`). The `--project` flag installs into the repository rather than the home directory.

How does an agent learn the command list?

With `pup --help` and the machine-readable `pup agent schema`; the full reference is in `docs/COMMANDS.md`.

What about profiling?

Pup does not cover it yet: use the Datadog MCP server with the `profiling` toolset from the Datadog documentation.

Discussion

No questions or reviews yet.

CLI tools
Editors’ pick

Official Datadog skills: monitors, logs, APM, cloud integrations, RUM and LLM observability from inside the agent

SkillBroad access185
Official

Datadog's official managed MCP server: logs, metrics, traces and incidents right in your agent

MCP serverExtended access48

Independent Datadog MCP server: an agent reads incidents, monitors, logs, metrics, traces and RUM and schedules downtimes.

MCP serverExtended access143

A self-hosted control center for coding agents: Claude Code, Codex, Copilot, OpenCode and Pi from your phone, desktop and terminal

CLIExtended access23.9K

Official Dynatrace skills for agents: DQL, traces, logs, Kubernetes, cloud and security work inside Claude Code and Cursor.

SkillExtended access179

Official Dynatrace CLI with kubectl-style syntax and a bundled skill: workflows, dashboards, DQL queries and SLOs

CLIExtended access204

README in short

A Rust program with builds for macOS, Linux and Windows, an npm package `@datadog/pup` and a skill set in `SKILL.md`. The command reference is in `docs/COMMANDS.md` and setup goes through `pup auth login`.

README badge

Report a problem