Usage Examples

Common workflows and usage patterns for Pup CLI.

Authentication

OAuth2 Login (Recommended)

# Login with default site (datadoghq.com)
pup auth login

# Login with specific site
pup --site=datadoghq.eu auth login

# Check authentication status
pup auth status

# Logout
pup auth logout

API Key Authentication (Legacy)

export DD_API_KEY="your-api-key"
export DD_APP_KEY="your-app-key"
export DD_SITE="datadoghq.com"

Metrics

List Metrics

# List all metrics
pup metrics list

# Filter by pattern
pup metrics list --filter="system.*"
pup metrics list --filter="custom.app.*"

Search Metrics (v1 API)

# Classic query syntax
pup metrics search --query="avg:system.cpu.user{*}" --from="1h"

# Search with aggregation and grouping
pup metrics search --query="sum:app.requests{env:prod} by {service}" --from="4h"

Query Metrics (v2 API)

# Timeseries formula query
pup metrics query --query="avg:system.cpu.user{*}" --from="1h" --to="now"

# Query with aggregation
pup metrics query --query="sum:app.requests{env:prod} by {service}" --from="4h"

Monitors

List Monitors

# List all monitors
pup monitors list

# Filter by tag
pup monitors list --tag="env:production"
pup monitors list --tag="team:backend"

# Multiple tags
pup monitors list --tag="env:prod" --tag="service:api"

Get Monitor Details

# Get specific monitor by ID
pup monitors get 12345678

Delete Monitor

# Delete monitor (prompts for confirmation)
pup monitors delete 12345678

# Skip confirmation
pup monitors delete 12345678 --yes

Logs

Search Logs

# Search for errors in last hour
pup logs search --query="status:error" --from="1h" --to="now"

# Search by service
pup logs search --query="service:web-app status:warn" --from="30m"

# Complex query with attributes
pup logs search --query="@user.id:12345 status:error" --limit=100

# Search with time range
pup logs search \
  --query="service:api" \
  --from="2024-02-04T10:00:00Z" \
  --to="2024-02-04T11:00:00Z"

Aggregate Logs

# Count logs by status
pup logs aggregate \
  --query="service:web-app" \
  --from="1h" \
  --compute="count" \
  --group-by="status"

# Average duration by service
pup logs aggregate \
  --query="service:web-app" \
  --from="1h" \
  --compute="avg(@duration)" \
  --group-by="service"

# 99th percentile latency by service
pup logs aggregate \
  --query="env:prod" \
  --from="30m" \
  --compute="percentile(@duration, 99)" \
  --group-by="service"

# Multiple metrics in one query (comma-separated)
pup logs aggregate \
  --query="service:web-app" \
  --from="1h" \
  --compute="count,avg(@duration),percentile(@duration, 95)" \
  --group-by="service,status"

Search Logs in Specific Storage Tier

# Search Flex logs (cost-optimized storage tier)
pup logs search --query="service:api" --from="7d" --storage="flex"

# Search online archives (long-term storage)
pup logs search --query="status:error" --from="30d" --storage="online-archives"

# Search standard indexes (default, fastest tier)
pup logs search --query="service:web-app" --from="1h" --storage="indexes"

# Use Datadog's default storage behavior
pup logs search --query="status:warn" --from="1h"

Dashboards

List Dashboards

# List all dashboards
pup dashboards list

# Output as table
pup dashboards list --output=table

Get Dashboard

# Get dashboard details
pup dashboards get "abc-123-def"

# Get public URL for sharing
pup dashboards url "abc-123-def"

# Open with a live 1 week time window
pup dashboards url "abc-123-def" --from=now-1w --to=now --live=true

Delete Dashboard

pup dashboards delete "abc-123-def" --yes

Database Monitoring

Search DBM Query Samples

# Search recent DBM activity samples
pup dbm samples search --query="dbm_type:activity service:orders env:prod" --from="1h" --limit=10

# Search a specific database host
pup dbm samples search --query="db.hostname:primary-db service:checkout" --from="30m" --sort="asc"

# Search an explicit time window
pup dbm samples search \
  --query="service:payments @db.statement_type:select" \
  --from="2024-02-04T10:00:00Z" \
  --to="2024-02-04T11:00:00Z" \
  --limit=25

SLOs

List SLOs

# List all SLOs
pup slos list

# Filter by name or API-supported query string
pup slos list --query="monitor-history-reader"

# Filter by a single SLO tag
pup slos list --tags-query="team:slo-app"

# Filter by metric query and paginate results
pup slos list --metrics-query="sum:requests.error{service:api}" --limit=25 --offset=50

Get SLO

pup slos get "abc-123-def"

Create SLO

pup slos create \
  --name="API Availability" \
  --type="metric" \
  --target=99.9 \
  --timeframe="30d"

Manage SLO Corrections

# List corrections for SLO
pup slos corrections list "abc-123-def"

# Create correction
pup slos corrections create "abc-123-def" \
  --start="2024-02-04T10:00:00Z" \
  --end="2024-02-04T11:00:00Z" \
  --category="deployment"

Incidents

List Incidents

# List all incidents
pup incidents list

# Filter by status
pup incidents list --status="active"

Get Incident

pup incidents get "abc-123"

Create Incident

pup incidents create \
  --title="High Error Rate in API" \
  --severity="SEV-2" \
  --customer-impacted=true

Update Incident

pup incidents update "abc-123" --status="resolved"

RUM (Real User Monitoring)

List RUM Applications

pup rum apps list

Get RUM Application

pup rum apps get "abc-123"

Search RUM Sessions

pup rum sessions search \
  --query="@application.id:abc-123" \
  --from="1h"

Security

List Security Rules

pup security rules list

Get Security Rule

pup security rules get "abc-123"

List Security Signals

pup security signals list --from="1h"

Search Security Findings

pup security findings search \
  --query="@severity:high"

APM Service Config

Get Service Instance Configuration

# Get service instance metadata (IDs, hostnames, config IDs)
pup apm service-config get --service-name my-service

# Filter by environment
pup apm service-config get --service-name my-service --env prod

# Filter by specific instance IDs
pup apm service-config get --service-name my-service --service-instance-ids "id-1,id-2"

APM Service Library Config

Get Service Library Configuration

# Get tracer configuration for a service across all instances
pup apm service-library-config get --service-name my-service

# Filter by environment
pup apm service-library-config get --service-name my-service --env prod

# Filter by language
pup apm service-library-config get --service-name my-service --env prod --language python

# Only show configs where instances disagree (useful for finding config drift)
pup apm service-library-config get --service-name my-service --mixed

APM Troubleshooting

List Instrumentation Errors for a Host

# Show APM instrumentation errors for a specific host
pup apm troubleshooting list --hostname my-host

# Narrow results to a specific time window
pup apm troubleshooting list --hostname my-host --timeframe 4h

Fleet Tracers

List Tracers Across the Fleet

# List all tracers (telemetry-derived service names, language, runtime IDs)
pup fleet tracers list

# Filter tracers by environment
pup fleet tracers list --filter "env:prod"

# Filter tracers by hostname
pup fleet tracers list --filter "hostname:my-host"

# Paginate results
pup fleet tracers list --filter "env:prod" --page-size 50 --page-number 0

List Tracers for a Specific Agent

# Get tracers running on a specific agent
pup fleet agents tracers <agent-key>

# With pagination
pup fleet agents tracers <agent-key> --page-size 20

List Instrumented Pods (K8s)

# List pods instrumented by SSI in a cluster
pup fleet instrumented-pods list <cluster-name>

Fleet Clusters

List Kubernetes Clusters in the Fleet

# List all clusters
pup fleet clusters list

# Filter by cluster name
pup fleet clusters list --filter "cluster_name:production"

# Paginate
pup fleet clusters list --filter "env:prod" --page-size 50 --page-number 0

Live Debugger

Service Context

# Get full service context as JSON (environments, instances, probe support)
pup debugger context my-service

# Filter to a specific environment
pup debugger context my-service --env staging

# Select specific fields for compact output: service, language, envs, repo
pup debugger context my-service --fields service,language,envs

List Log Probes

# List all log probes
pup debugger probes list

# Filter by service
pup debugger probes list --service my-service

Get Probe Details

pup debugger probes get "probe-id"

Create a Log Probe

# Create a probe with capture expressions (recommended)
pup debugger probes create \
  --service my-service \
  --env staging \
  --probe-location com.example.MyClass:myMethod \
  --capture "request.id" --capture "user.name"

# Use a method signature from --view probe-locations output
pup debugger probes create \
  --service my-service \
  --env staging \
  --probe-location "com.example.MyClass:myMethod(String, int)" \
  --capture "request.id"

# Increase capture depth for nested objects (default: 1)
pup debugger probes create \
  --service my-service \
  --env staging \
  --probe-location com.example.MyClass:myMethod \
  --capture "response.body" --depth 3

# Create with full snapshot capture
pup debugger probes create \
  --service my-service \
  --env staging \
  --probe-location com.example.MyClass:myMethod \
  --capture

# Create with a custom template
pup debugger probes create \
  --service my-service \
  --env staging \
  --probe-location com.example.MyClass:myMethod \
  --capture "userId" \
  --template "User {userId} called, took {@duration}ms"

# Create with a condition
pup debugger probes create \
  --service my-service \
  --env staging \
  --probe-location com.example.MyClass:myMethod \
  --capture "userId" \
  --condition "userId != null"

# Create with custom budget and TTL
pup debugger probes create \
  --service my-service \
  --env staging \
  --probe-location com.example.MyClass:myMethod \
  --budget 500 --ttl 2h

Delete a Log Probe

pup debugger probes delete "probe-id"

Watch Probe Events

# Stream probe events — compact output (message, captures, timestamp)
pup debugger probes watch "probe-id" --fields "message,captures,timestamp" --wait 10

# Template message only (one line per event)
pup debugger probes watch "probe-id" --fields "message" --limit 5

# Full debugger payload (default — trimmed to just the debugger field)
pup debugger probes watch "probe-id"

# Limit to 5 events
pup debugger probes watch "probe-id" --limit 5

# Custom timeout
pup debugger probes watch "probe-id" --timeout 300

# Start from a specific time
pup debugger probes watch "probe-id" --from 1h

# Wait for a newly created probe to become available (useful in pipelines)
pup debugger probes watch "probe-id" --wait 30

Pipeline: Create and Watch

# Search for a method, create a probe, and watch events
# Note: --view probe-locations may output signatures like TYPE:METHOD(args)
pup symdb search --service my-service --query MyController --view probe-locations \
  | head -1 \
  | xargs -I{} pup debugger probes create --service my-service --env staging --probe-location {} --capture --ttl 1h \
  | jq -r .data.id \
  | xargs -I{} pup debugger probes watch {} --fields "message,captures,timestamp" --wait 30 --limit 5

SymDB (Symbol Database)

Search Scopes

# Search for classes/methods by name
pup symdb search --service my-service --query MyController

# List all scopes in a service
pup symdb search --service my-service

# Filter by service version
pup symdb search --service my-service --query MyController --version 1.2.3

Output Views

# Full JSON response (default)
pup symdb search --service my-service --query MyController --view full

# Scope names only
pup symdb search --service my-service --query MyController --view names

# Probe locations (type:method or type:method(args) format)
pup symdb search --service my-service --query MyController --view probe-locations

Containers

List Containers

# List all containers
pup containers list

# Filter by tags
pup containers list --filter-tags="env:production"

# Group by image
pup containers list --group-by="image_name"

List Container Images

# List all container images
pup containers images list

# Filter images by tags
pup containers images list --filter-tags="env:production"

Infrastructure

List Hosts

# List all hosts
pup infrastructure hosts list

# Filter by tag
pup infrastructure hosts list --filter="env:production"

Fleet

List Fleet Agents

# Filter agents by hostname
pup fleet agents list --filter "hostname:my-host"

# Filter by IP address
pup fleet agents list --filter "ip_address:1.2.3.4"

# Boolean filter expression
pup fleet agents list --filter "(hostname:host-a OR hostname:host-b) AND env:prod"

Get Host

pup infrastructure hosts get "host-name"

Users & Organizations

List Users

pup users list

Get User

pup users get "user-id"

List Roles

pup users roles list

Get Organization

pup organizations get

API Keys

List API Keys

pup api-keys list

Get API Key

pup api-keys get "key-id"

Create API Key

pup api-keys create --name="CI/CD Key"

Delete API Key

pup api-keys delete "key-id" --yes

Synthetics

List Synthetic Tests

pup synthetics tests list

Get Synthetic Test

pup synthetics tests get "test-id"

List Synthetic Locations

pup synthetics locations list

Workflows

Get a Workflow

pup workflows get <workflow-id>

Create a Workflow

pup workflows create --file=workflow.json

Update a Workflow

pup workflows update <workflow-id> --file=workflow.json

Delete a Workflow

pup workflows delete <workflow-id>

Execute a Workflow

# Run with inline payload (requires DD_API_KEY + DD_APP_KEY)
pup workflows run <workflow-id> --payload '{"key": "value"}'

# Run with payload from file
pup workflows run <workflow-id> --payload-file=params.json

# Run and wait for completion (default timeout: 5m)
pup workflows run <workflow-id> --wait

# Run with custom timeout
pup workflows run <workflow-id> --wait --timeout 2m

Manage Workflow Instances

# List recent executions
pup workflows instances list <workflow-id>

# List with pagination
pup workflows instances list <workflow-id> --limit=20 --page=2

# Get instance details
pup workflows instances get <workflow-id> <instance-id>

# Cancel a running instance
pup workflows instances cancel <workflow-id> <instance-id>

Runbooks

Runbooks are YAML files stored in ~/.config/pup/runbooks/ that encode multi-step operational procedures. Each step runs a pup command, shell command, HTTP request, Datadog Workflow, or an interactive confirmation.

List Available Runbooks

pup runbooks list
pup runbooks list --tag=type:deployment

Inspect a Runbook

pup runbooks describe deploy-service

Run a Runbook

# Run with required variables via --arg flags
pup runbooks run deploy-service --arg SERVICE=payments --arg VERSION=1.2.3

# Dry-run (print steps without executing)
pup runbooks run incident-triage --dry-run

Import a Runbook

# Copy a runbook YAML into ~/.config/pup/runbooks/
pup runbooks import ./my-runbook.yaml

Validate a Runbook

pup runbooks validate ./my-runbook.yaml

Example Runbook (YAML)

name: restart-service
description: Safely restart a service after checking monitors
vars:
  SERVICE:
    description: Service name
    required: true

steps:
  - name: Check active monitors
    kind: pup
    run: monitors list --tags="service:{{SERVICE}}"
    capture: MONITORS_JSON

  - name: Confirm restart
    kind: confirm
    message: "Restart {{SERVICE}}? Review monitors above."

  - name: Trigger restart workflow
    kind: datadog-workflow
    workflow_id: "abc-123"
    inputs:
      service: "{{SERVICE}}"
    on_failure: fail

  - name: Notify via webhook
    kind: http
    url: https://hooks.example.com/notify
    method: POST
    body: '{"text": "Restarted {{SERVICE}}"}'
    content_type: application/json
    on_failure: warn

Reusable Step Templates

Store shared step logic in ~/.config/pup/runbooks/_templates/<name>.yaml:

# _templates/slack-notify.yaml
kind: http
url: "{{SLACK_WEBHOOK}}"
method: POST
body: '{"text": "{{MESSAGE}}"}'
content_type: application/json
on_failure: warn

Reference a template in any runbook step:

steps:
  - name: Notify Slack
    template: slack-notify
    # Override any template field as needed

See docs/examples/runbooks/ for complete examples.

IDP (Service Catalog)

Get Full Service Context

# Get owner, on-call, health, dependencies, and metadata gaps in one call
pup idp assist my-service

# Useful as a starting point for incident response or code review
pup idp assist payments-api

Find Entities

# Search services by name (fuzzy match)
pup idp find payments

# Use kind: prefix to search other entity types
pup idp find "kind:team AND name:backend"

Get Ownership and On-Call

# Show owning team and current on-call responders
pup idp owner my-service

Show Service Dependencies

# List upstream (callers) and downstream (callees) services
pup idp deps my-service

Register a Service Definition

# POST a service.datadog.yaml to the Service Definitions API
pup idp register service.datadog.yaml

# Verify after registration
pup idp assist my-service

Incident Response Workflow with IDP

# Get full service context immediately
pup idp assist payments-api

# Investigate alerts for the service
pup monitors list --tag="service:payments-api"

# Check who is on-call
pup idp owner payments-api

# Review upstream services that may be affected
pup idp deps payments-api

Costs

Datadog Cost (Usage Metering)

# Projected cost for current month
pup costs datadog projected

# Cost breakdown by org for a date range
pup costs datadog by-org --start-month="2024-01-01T00:00:00Z" --end-month="2024-03-01T00:00:00Z"

# Monthly cost attribution
pup costs datadog attribution --start="2024-01-01T00:00:00Z"

# AWS CUR config management
pup costs datadog aws-config list
pup costs datadog aws-config get 42
pup costs datadog aws-config create --file=aws-cur.json
pup costs datadog aws-config delete 42

# Azure UC config management
pup costs datadog azure-config list

# GCP usage cost config management
pup costs datadog gcp-config list

Cloud Cost Management (CCM)

Custom Costs

# List uploaded custom cost files
pup costs ccm custom-costs list

# Filter by status or sort
pup costs ccm custom-costs list --status=SUCCESS --sort=created_at

# Get a specific custom cost file
pup costs ccm custom-costs get <file-id>

# Upload a custom cost CSV (multipart PUT)
pup costs ccm custom-costs upload --file=costs.csv

# Upload with a schema version
pup costs ccm custom-costs upload --file=costs.csv --version=2

# Delete a custom cost file
pup costs ccm custom-costs delete <file-id>

Tag Descriptions

# List all tag descriptions
pup costs ccm tag-descriptions list

# Filter by cloud provider
pup costs ccm tag-descriptions list --cloud=aws

# Get description for a specific tag key
pup costs ccm tag-descriptions get --tag-key=env

# Generate an AI-suggested description for a tag
pup costs ccm tag-descriptions generate --tag-key=team

# Create or update a tag description
pup costs ccm tag-descriptions upsert --tag-key=env --description="Deployment environment"

# Delete a tag description
pup costs ccm tag-descriptions delete --tag-key=env

Tag Metadata

# List all tag metadata for a month
pup costs ccm tag-metadata list --month=2024-01

# Filter by provider and metric
pup costs ccm tag-metadata list --month=2024-01 --provider=aws --metric=aws.cost

# List tag sources
pup costs ccm tag-metadata tag-sources --month=2024-01

# List available metrics
pup costs ccm tag-metadata metrics --month=2024-01

# List orchestrators
pup costs ccm tag-metadata orchestrators --month=2024-01

# List supported currencies
pup costs ccm tag-metadata currency --month=2024-01

Tags and Tag Keys

# List all CCM tags
pup costs ccm tags list

# Filter by metric
pup costs ccm tags list --metric=aws.cost

# List tag keys
pup costs ccm tag-keys list

# Get details for a specific tag key
pup costs ccm tag-keys get env

Budgets

# List all budgets
pup costs ccm budgets list

# Get a specific budget
pup costs ccm budgets get <budget-id>

# Get with actual and forecasted spend for a time range
pup costs ccm budgets get <budget-id> \
  --start="2024-01-01T00:00:00Z" \
  --end="2024-03-01T00:00:00Z" \
  --actual --forecast

# Create or update a budget (upsert)
pup costs ccm budgets upsert --file=budget.json

# Delete a budget
pup costs ccm budgets delete <budget-id>

# Validate a budget payload without saving
pup costs ccm budgets validate --file=budget.json

Commitment Programs (Reserved Instances & Savings Plans)

# Utilization of RIs over a time range
pup costs ccm commitments utilization \
  --provider=aws --product=EC2 \
  --from=2024-01-01T00:00:00Z --to=2024-02-01T00:00:00Z

# Coverage by Savings Plans
pup costs ccm commitments coverage \
  --provider=aws --product=EC2 \
  --from=2024-01-01T00:00:00Z --to=2024-02-01T00:00:00Z \
  --commitment-type=SP

# Savings summary
pup costs ccm commitments savings \
  --provider=aws --product=RDS \
  --from=2024-01-01T00:00:00Z --to=2024-02-01T00:00:00Z

# On-demand hot spots (unreserved spend)
pup costs ccm commitments hotspots \
  --provider=aws --product=EC2 \
  --from=2024-01-01T00:00:00Z --to=2024-02-01T00:00:00Z

# Utilization time series
pup costs ccm commitments utilization-ts \
  --provider=aws --product=EC2 \
  --from=2024-01-01T00:00:00Z --to=2024-02-01T00:00:00Z

# Coverage time series
pup costs ccm commitments coverage-ts \
  --provider=aws --product=EC2 \
  --from=2024-01-01T00:00:00Z --to=2024-02-01T00:00:00Z

# Savings time series
pup costs ccm commitments savings-ts \
  --provider=aws --product=EC2 \
  --from=2024-01-01T00:00:00Z --to=2024-02-01T00:00:00Z

# List all commitment programs
pup costs ccm commitments list \
  --provider=aws --product=EC2 \
  --from=2024-01-01T00:00:00Z --to=2024-02-01T00:00:00Z

Output Formatting

JSON Output (Default)

pup monitors list --output=json

YAML Output

pup monitors list --output=yaml

Table Output

pup monitors list --output=table

Custom Fields

pup monitors list --fields="id,name,type,status"

Advanced Usage

Custom Config File

pup --config=/path/to/config.yaml monitors list

Specify Datadog Site

pup --site=datadoghq.eu monitors list

Verbose Output (Debug)

pup --verbose monitors list

Skip Confirmation Prompts

pup --yes monitors delete 12345678

Read-Only Mode

# Block all write operations (create, update, delete)
pup --read-only monitors list
pup --read-only dashboards list

# Also available via env var or config file
DD_READ_ONLY=true pup monitors list

Common Workflows

Monitoring Dashboard

# List monitors for a service
pup monitors list --tag="service:api" --output=table

# Check recent logs
pup logs search --query="service:api" --from="1h" --output=table

# Query metrics
pup metrics query --query="avg:api.latency{*}" --from="1h"

Incident Response

# Create incident
pup incidents create --title="API Down" --severity="SEV-1"

# Search related logs
pup logs search --query="status:error service:api" --from="1h"

# Check monitors
pup monitors list --tag="service:api"

# Update incident status
pup incidents update "incident-id" --status="investigating"

Security Audit

# List recent security signals
pup security signals list --from="24h"

# Check security rules
pup security rules list

# Search security findings
pup security findings search --query="@severity:critical"

# Review audit logs
pup audit-logs list --from="7d"

Infrastructure Review

# List all hosts
pup infrastructure hosts list --output=table

# Get host details
pup infrastructure hosts get "host-name"

# Review host tags
pup tags list --output=table

Time Range Formats

Relative Times

--from="1h"    # 1 hour ago
--from="30m"   # 30 minutes ago
--from="7d"    # 7 days ago
--from="now"   # Current time

Absolute Times

--from="2024-02-04T10:00:00Z"
--to="2024-02-04T11:00:00Z"

Unix Timestamps

--from="1707048000"  # Unix timestamp in seconds

Environment Variables

# Authentication
export DD_API_KEY="your-api-key"
export DD_APP_KEY="your-app-key"
export DD_SITE="datadoghq.com"

# Configuration
export PUP_CONFIG="/path/to/config.yaml"
export PUP_OUTPUT="json"
export PUP_LOG_LEVEL="debug"

ACP Server (AI Agent Integration)

pup acp serve starts a local HTTP server that lets AI coding assistants and agents talk directly to Datadog Bits AI. It speaks two protocols:

ACP (Agent Communication Protocol) — for ACP-native clients
OpenAI-compatible — for tools like opencode, Cursor, or any @ai-sdk/openai-compatible client

Quick Start

# Authenticate first (notebooks_read + notebooks_write scopes required)
pup auth login

# Start the server (auto-discovers your first Datadog Bits AI agent)
pup acp serve

# Specify a particular agent
pup acp serve --agent-id <uuid>

# Custom port or bind address
pup acp serve --port 8080
pup acp serve --host 0.0.0.0 --port 9099

Endpoints

Method	Path	Protocol	Description
GET	`/agent.json`	ACP	Agent card / capability discovery
POST	`/runs`	ACP	Synchronous run — returns full response
POST	`/runs/stream`	ACP	Streaming run — SSE events
GET	`/models` or `/v1/models`	OpenAI	Model list
POST	`/chat/completions` or `/v1/chat/completions`	OpenAI	Chat completions (streaming or sync)

Testing with curl

# ACP sync
curl -s -X POST http://127.0.0.1:9099/runs \
  -H "Content-Type: application/json" \
  -d '{"input": [{"role": "user", "content": [{"type": "text", "text": "list my monitors with status alert"}]}]}' \
  | jq .output[0].content[0].text

# ACP streaming
curl -X POST http://127.0.0.1:9099/runs/stream \
  -H "Content-Type: application/json" \
  -d '{"input": [{"role": "user", "content": [{"type": "text", "text": "what services have errors in the last hour?"}]}]}'

# OpenAI-compatible
curl -s -X POST http://127.0.0.1:9099/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model": "datadog-ai", "messages": [{"role": "user", "content": "how many monitors are currently alerting?"}]}' \
  | jq .choices[0].message.content

opencode Setup

Add to ~/Library/Application Support/opencode/opencode.jsonc (macOS) or ~/.config/opencode/opencode.jsonc (Linux):

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "datadog": {
      "name": "Datadog AI",
      "npm": "@ai-sdk/openai-compatible",
      "models": {
        "datadog-ai": {
          "name": "Datadog AI Agent"
        }
      },
      "options": {
        "baseURL": "http://127.0.0.1:9099"
      }
    }
  }
}

Then start the server (pup acp serve) and select the Datadog AI provider in opencode.

Configuration File

Create ~/.config/pup/config.yaml:

site: datadoghq.com
output: json
verbose: false

# Default time ranges
default_from: 1h
default_to: now

# Output preferences
output_format: json
table_max_width: 120

FilesExpand file tree

EXAMPLES.md

Latest commit

History