Skip to content

Latest commit

 

History

History
510 lines (419 loc) · 31.9 KB

COMMANDS.md

File metadata and controls

510 lines (419 loc) · 31.9 KB

Command Reference

Complete reference for the command groups in Pup. Run pup --help (or pup agent schema --compact) for the live, exhaustive list as built into your binary — the table below is curated documentation and may lag the source.

Command Pattern

pup <domain> <action> [options]           # Simple commands
pup <domain> <subgroup> <action> [options] # Nested commands

Status Legend

  • ✅ WORKING - Command compiles and runs (requires valid auth)
  • ⚠️ API BLOCKED - Implementation correct, waiting for API client library updates
  • ⏳ PLACEHOLDER - Skeleton implementation, API endpoints pending

Command Index

Domain Subcommands File Status
acp serve src/commands/acp.rs ✅
auth login, logout, status, token, refresh src/commands/auth.rs ✅
metrics query, list, search, timeseries, metadata, tags, submit src/commands/metrics.rs ✅
logs search, list, aggregate, patterns, saved-views (list, get, create, delete) src/commands/logs.rs ✅
traces get, search, aggregate, metrics (list, get, create, update, delete) src/commands/traces.rs ✅
monitors list, get, create, update, delete, search, diff src/commands/monitors.rs ✅
dashboards list, get, create, update, diff, delete, url, annotations (list, get-page, create, update, delete) src/commands/dashboards.rs, src/commands/annotations.rs ✅
snapshots create src/commands/snapshots.rs ✅
dbm samples (search) src/commands/dbm.rs ✅
ddsql table, spec, schema (tables, columns) src/commands/ddsql.rs ✅
debugger probes (list, get, create, delete, watch) src/commands/debugger.rs ✅
slos list, get, create, update, diff, delete, status src/commands/slos.rs ✅
incidents list, get, attachments, settings, handles, postmortem-templates src/commands/incidents.rs ✅
rum apps, metrics, retention-filters, sessions, events, aggregate, playlists, replay, viewership, heatmaps src/commands/rum.rs ✅
cicd pipelines, events, tests, dora, flaky-tests src/commands/cicd.rs ✅
static-analysis custom-rulesets (get, update, delete), custom-rules (get, create, delete, revisions, revision) src/commands/static_analysis.rs ✅
downtime list, get, cancel src/commands/downtime.rs ✅
tags list, get, add, update, delete src/commands/tags.rs ✅
events post, list, search, get src/commands/events.rs ✅
on-call teams (CRUD, memberships), pages (newest-first list, get, create) src/commands/on_call.rs ✅
audit-logs list, search src/commands/audit_logs.rs ✅
api-keys list, get, create, delete src/commands/api_keys.rs ✅
app-keys list, get, create, update, delete src/commands/app_keys.rs ✅
infrastructure hosts (list, get) src/commands/infrastructure.rs ✅
synthetics tests, locations, suites, downtime src/commands/synthetics.rs ✅
symdb search src/commands/symdb.rs ✅
logs-restriction list, get, create, update, delete, roles (list, add) src/commands/logs_restriction.rs ✅
processes list src/commands/processes.rs ✅
users list, get, roles, service-accounts (create, app-keys CRUD) src/commands/users.rs ✅
notebooks list, get, create, update, edit, diff, delete (get/create/update/edit accept --markdown), annotations (list, get-page, create, update, delete) src/commands/notebooks.rs, src/commands/annotations.rs ✅
security rules, signals, findings, content-packs, risk-scores src/commands/security.rs ✅
organizations get, list src/commands/organizations.rs ✅
service-catalog list, get src/commands/service_catalog.rs ✅
idp kinds (list, describe), entities (query), assist, find, owner, deps, register, migrate-schema src/commands/idp/ ✅
error-tracking issues (search, get) src/commands/error_tracking.rs ✅
scorecards rules (list, create, update, delete), outcomes (list, batch-create) src/commands/scorecards.rs ✅
usage summary, hourly src/commands/usage.rs ✅
apm services (list, stats, operations, resources), entities (list), dependencies (list), flow-map, troubleshooting (list), service-config (get), service-library-config (get) src/commands/apm.rs ✅
containers list, images (list) src/commands/containers.rs ✅
costs datadog (projected, attribution, by-org, aws-config, azure-config, gcp-config), ccm (custom-costs, tag-descriptions, tag-metadata, tags, tag-keys, budgets, commitments) src/commands/cost.rs, src/commands/cost_ccm.rs ✅
product-analytics events send src/commands/product_analytics.rs ✅
experiments list, get, create, update, delete, start, conclude, cancel, diagnostics, traffic-summary, results, refresh-results, analysis-plan (get, update), metric-groups (list, create, create-from-collection, update, delete), metrics (list, get, create, update, delete), metric-collections (list, get, create, update, delete), subject-types (list, get, create, update, delete, set-default), protocols (list, get) src/commands/experiments.rs ✅
profiling profiles (list, download — design-partner orgs only), services (list), profile-types (list), explore (flamegraph, callgraph, timeline) src/commands/profiling.rs ✅
datasets list, get, create, update, delete src/commands/datasets.rs ✅
data-deletion requests (list, create, cancel) src/commands/data_deletion.rs ✅
data-governance scanner-rules (list) src/commands/data_governance.rs ✅
obs-pipelines list, get, create, update, diff, delete, validate src/commands/obs_pipelines.rs ✅
llm-obs projects (create, list), experiments (create, list, update, delete, summary, events (list, get, submit), metric-values, dimension-values), datasets (create, list, batch-update, clone, restore, records, records-add, records-all, records-full), spans (search), patterns (configs (list, get, create, update), runs (list, status), topics, topics-with-points, points), agent-insights (list, get, update-status, submit-feedback), annotation-queues (create, list, update, delete, interactions (add, delete, list), schema (get, update), annotations (upsert, delete)), model-pricing src/commands/llm_obs.rs ✅
reference-tables list, get, create, batch-query src/commands/reference_tables.rs ✅
network flows list, devices (list, get, interfaces, tags), interfaces (list, update) src/commands/network.rs ✅
cloud aws, gcp, azure, oci src/commands/cloud.rs ✅
integrations list (--configured, --available, --broken, --manageable, --search), slack, pagerduty, webhooks, jira, servicenow, google-chat, ms-teams, okta, cloudflare, databricks, elastic-cloud, snowflake, twilio, confluent-cloud, fastly, statuspage src/commands/integrations.rs ✅
misc ip-ranges, status src/commands/misc.rs ✅
cases create, get, search, count, aggregate, assign, unassign, archive, update-*, comments, watchers, links, views, types, custom-attributes, projects (automation-rules, notification-rules, favorites), jira, servicenow, incidents, notebooks, insights, move src/commands/cases.rs ✅
status-pages pages, components, degradations src/commands/status_pages.rs ✅
code-coverage branch-summary, commit-summary src/commands/code_coverage.rs ✅
hamr connections (get, create) src/commands/hamr.rs ✅
fleet agents (list, get, versions, tracers), deployments (list, get, configure, upgrade, cancel), schedules (list, get, create, update, delete, trigger), tracers (list), clusters (list), instrumented-pods (list) src/commands/fleet.rs ✅
skills list, install, path (positional <platform>: claude/cursor/codex/opencode/windsurf/gemini/pi/devin/hermes/all; --name, --type, --project for project-local scope; bundled skills include supplementary references) src/commands/skills.rs ✅
runbooks list, describe, run, import, validate src/commands/runbooks.rs ✅
setup (all arguments passed to AI Setup, run from npm with npx; needs Node.js 22+) src/commands/setup.rs ✅
workflows get, create, update, diff, delete, run, instances (list, get, cancel), connections (get, create, update, delete) src/commands/workflows.rs ✅
investigations list, get, trigger src/commands/investigations.rs ✅
change-requests create, get, update, create-branch, decisions (update, delete) src/commands/change_management.rs ✅
change-stories list src/commands/change_stories.rs ✅
app-builder list, get, create, update, delete, delete-batch, publish, unpublish src/commands/app_builder.rs ✅
governance insights (list), controls (list, get), detections (list, get, update, mitigate), limits (list), best-practices (list), tag-rules (list, get, create, update, delete, score) src/commands/governance.rs, src/commands/tag_rules.rs ✅

Note: RUM command is fully operational. Apps and sessions work completely. Metrics and retention-filters support list/get operations (create/update/delete operations pending due to complex API type structures).

Auth note: All workflow commands require DD_API_KEY + DD_APP_KEY. OAuth2 bearer tokens are not supported for workflow operations.

Common Patterns

List Operations

pup <domain> list [--flags]
pup monitors list --tags="env:production"
pup dashboards list

Get Operations

pup <domain> get <id>
pup monitors get 12345678
pup slos get abc-123-def

Full Trace Retrieval

traces get uses the preview GET /api/v2/trace/{trace_id} endpoint rather than indexed span search. It requires apm_read and accepts a 32-character hex ID or up to 39 decimal digits. To save the full response without reading it into agent context, redirect stdout:

pup --read-only traces get <TRACE_ID> > trace.json

Spans are under .data.attributes.spans; check .data.attributes.is_truncated before treating the response as complete. The command requests all available fields. Preserve integer ID precision when parsing the saved JSON.

traces get --pruned uses the preview GET /api/v2/pruned_trace/{trace_id} endpoint, which returns a summarized span tree instead of the full span list. It also requires apm_read. The tree is under .data.attributes.summarized_trace.root, and the pruning flags below are rejected unless --pruned is set. Each node's hidden_child_spans_count shows how many child spans were pruned.

pup traces get <TRACE_ID> --pruned --only-service-entry-spans
pup traces get <TRACE_ID> --pruned --expand-span-id=<SPAN_ID>
pup traces get <TRACE_ID> --pruned --include-path="service:api" --tag-exclude="^_dd\."

Search/Query

pup logs search --query="status:error" --from="1h"
pup logs search --query="service:api" --from="7d" --storage="flex"
pup logs patterns --query="status:error" --pattern-field="message" --from="1h"
pup logs query --query="service:api" --index="main,security" --from="1h"
pup logs saved-views create --file=saved-view.json
pup dbm samples search --query="dbm_type:activity service:orders env:prod" --from="1h" --limit=10
pup metrics search --query="avg:system.cpu.user{*}" --from="1h"
pup metrics query --query="avg:system.cpu.user{*}" --from="1h"
pup metrics tags list system.cpu.user --window-seconds=3600
pup metrics timeseries --file=request.json
pup events search --query="@user.id:12345"

IDP Entity Graph

Use kind discovery before writing flexible cross-entity queries:

# Curated kind index; add --all for the filtered live server inventory.
pup idp kinds list
pup idp kinds list --all --include-custom

# Live fields, relations, operators, examples, and caveats for one kind.
pup idp kinds describe service

# Query one result kind and optionally expand relations.
pup idp entities query 'kind:service AND owner:payments' \
  --field name,owner,contacts,service_health_status \
  --include owner_teams,systems

# Use scoped convenience wrappers when the kind is already known.
pup idp entities list --filter-kind=service \
  --field name,owner,contacts
pup idp entities search --filter-kind=service \
  --query='owner:idp OR team:idp'

# Continue an explicitly paginated query.
pup idp entities query 'kind:service' --cursor '<next_cursor>'

Every query must contain one unquoted kind:<kind> filter or a concrete ref:"ref:<kind>:<id>". Top-level OR across result kinds is rejected; group alternatives below a shared kind instead, for example kind:service AND (owner:idp OR team:idp). The scoped list and search commands build that kind scope from --filter-kind; search wraps --query in parentheses so OR alternatives cannot escape the kind. --field selects attributes and --include expands relations. Output is normalized and bounded for agents by default; pass --raw for the original JSON:API response.

Pagination

Paginated commands that have been migrated share two sets of flags:

  • Cursor style: --cursor and --page-size. Pass metadata.next_cursor back with --cursor.
  • Page style: --page (0-based) and --page-size. metadata.next_page is the value to pass to --page.

Hidden aliases keep older invocations working: --page-cursor, --page-token and --limit for cursor style; --per-page and --limit for page style. --page-number is also still accepted; its value goes to the API unchanged (so cases search --page-number 1 is still the first page). --offset is accepted too and converted to a page: --offset 50 --page-size 25 is --page 2. An offset that is not a multiple of the page size is an error that names the --page value to use. On ddsql schema tables/columns and llm-obs experiments events list, --offset must be an exact multiple of the page size.

--page-size must be at least 1 and at most the command's maximum; the default is the command's previous default. next_action in the output envelope names the command's real paging flag.

Command Style Default page size Max
deployment-gates gates list cursor server default 1000
cicd flaky-tests search cursor 100 none
llm-obs projects list cursor server default none
monitors search page 30 none
users list page 10 100
cases search page 10 none
ddsql schema tables / columns page 100 none
llm-obs experiments events list page 20 none

ddsql schema tables and columns now report metadata.next_page instead of metadata.next_offset. Other paginated commands keep their current flags until they are migrated.

Create/Update/Delete

pup <domain> create [--flags]
pup <domain> update <id> [--flags]
pup <domain> delete <id> [--yes]
pup events post --tags="version:1,application:web" --no_host --type=my_apps --aggregation_key=application:web --alert_type=info "Something big happened!" "And let me tell you all about it here!"

Nested Commands

pup rum apps list
pup rum metrics get <id>
pup cicd pipelines list
pup security rules list
pup infrastructure hosts list

Domain Categories

Data & Observability

  • metrics - Time-series metrics (query, list, get, search)
  • logs - Log search and analysis (search, list, aggregate, saved views)
  • dbm - Database Monitoring query samples (samples search)
  • traces - Full trace retrieval, span search/aggregation, and span-based metrics
  • rum - Real User Monitoring (apps, metrics, retention-filters, sessions)
  • events - Infrastructure events (post, list, search, get)
  • ddsql - DDSQL queries and discovery (table, spec, schema)
  • symdb - Symbol Database queries (search scopes, probe locations)

Monitoring & Alerting

  • monitors - Monitor management (list, get, delete)
  • dashboards - Dashboard management (list, get, delete, url)
  • slos - Service Level Objectives (list, get, delete, status)
  • synthetics - Synthetic monitoring (tests, locations, suites, downtime)
  • notebooks - Investigation notebooks (list, get, delete)
  • downtime - Monitor downtime (list, get, cancel)
  • status-pages - Status pages with components and degradations

Infrastructure & Performance

  • infrastructure - Host inventory (hosts list, hosts get)
  • network - Network monitoring (flows list, devices list/get/interfaces/tags, interfaces list/update)
  • tags - Host tag management (list, get, add, update, delete)
  • profiling - Continuous Profiler search/download (profiles list/download, services list, profile-types list, explore flamegraph/callgraph/timeline)
    • Uses unstable /api/unstable/profiling/pup/... endpoints with a reduced support guarantee: the latest pup release, plus 30 days for older versions (see EXAMPLES.md)
    • profiles list and profiles download are enabled only for design-partner organizations (Datadog-side enrollment); other orgs get HTTP 403 "not enabled for this organization" (exit code 403). The remaining profiling commands work for every org.

Security & Compliance

  • security - Security monitoring (rules, signals, findings, content-packs, risk-scores)
    • pup security findings mute --file <body.json> — Mute or unmute up to 100 findings (stable, SDK #1519/#1660)
    • pup security rules bulk-convert --file <payload.json> — Bulk convert existing rules to Terraform ZIP archive (SDK #1675)
  • static-analysis - Code security (custom-rulesets, custom-rules)
  • audit-logs - Audit trail (list, search)
  • data-governance - Sensitive data scanning (scanner-rules list)
  • governance - Governance Console insights, controls, detections, mitigations, limits, best practices, and tag rules. Detection writes support --dry-run and --yes. OAuth users opt in with pup auth login --extra-scopes governance_console_read; mitigations also need their declared permissions. WASM builds support reads and dry runs, but detection writes require a native build because the WASM transport cannot disable automatic redirects.

Cloud & Integrations

  • cloud - Cloud providers (aws, gcp, azure, oci)
  • integrations - Third-party integrations (list, slack, pagerduty, webhooks, jira, servicenow, google-chat, ms-teams, okta, cloudflare, databricks, elastic-cloud, snowflake, twilio, confluent-cloud, fastly, statuspage)
    • list shows the full catalog by default; --configured, --available, --broken, --manageable, and --search narrow it (the first three are mutually exclusive). Table columns are name, status (configured / available), the pup manage command, and health (ok / error) when a probe ran. Catalog IDs and full issue text are omitted from the table but kept in -o json. Health data backing --broken is only available for Amazon Web Services, Google Cloud, and Oracle Cloud Infrastructure.
    • Every pup integrations ... list renders as a table in a terminal; redirected/piped output and --jq stay JSON, and -o json forces JSON anywhere.
    • Account CRUD subcommands (okta, cloudflare, databricks, elastic-cloud, snowflake, twilio, confluent-cloud, fastly, statuspage) take request bodies via --file <json> so credentials stay out of shell history.
    • statuspage manages the Atlassian Statuspage integration; Datadog's own Status Pages product is pup status-pages.

Development & Quality

  • cicd - CI/CD visibility (pipelines, events, tests, dora, flaky-tests)
  • code-coverage - Code coverage summaries (branch, commit)
  • error-tracking - Error management (issues search, issues get); search supports --state, --team, --assignee filters
  • scorecards - Service quality (rules, outcomes)
  • service-catalog - Service registry (list, get)
  • idp - Entity graph discovery/query plus Service Catalog helpers (kinds, entities, assist, find, owner, deps, register, migrate-schema)
  • debugger - Live Debugger (probes list, get, create, delete, watch)

Operations & Incident Response

  • incidents - Incident management (list, get, attachments, settings, handles, postmortem-templates)
  • on-call - Team management (create, update, delete teams; manage memberships with roles) and pages (newest-first list, get, create)
  • cases - Case management (create, search, count, aggregate, assign, archive, update, comments, watchers, links, views, types, custom attributes, projects, automation rules, jira, servicenow, incidents, notebooks, move)
  • hamr - High Availability Multi-Region connections
  • fleet - Fleet Automation (agents, deployments, schedules, tracers, clusters, instrumented-pods)
  • runbooks - Local runbook execution engine (list, describe, run, import, validate)
  • workflows - Workflow Automation (get, create, update, diff, delete, run, instances, connections)
  • investigations - Bits AI SRE investigations (list, get, trigger)
  • change-requests - Change request management (create, get, update, create-branch, decisions)
  • change-stories - Change events for a service (deployments, feature flags, config, k8s, watchdog) over time window

Organization & Access

  • users - User management (list, get, roles)
  • organizations - Org settings (get, list)
  • api-keys - API key management (list, get, create, delete)
  • app-keys - Application key management (list, get, create, update, delete)

Cost & Usage

  • usage - Usage and billing (summary, hourly)
  • costs - Cost management: datadog subgroup (projected, attribution, by-org, aws-config, azure-config, gcp-config), ccm subgroup (custom-costs, tag-descriptions, tag-metadata, tags, tag-keys, budgets, commitments), and anomalies subgroup (list)

Configuration & Data Management

  • obs-pipelines - Observability pipelines (list, get, create, update, diff, delete, validate)
  • llm-obs - LLM Observability (projects, experiments, datasets, spans, agent insights, model pricing)
  • reference-tables - Reference tables for log enrichment (list, get, create, batch-query)
  • misc - Miscellaneous (ip-ranges, status)
  • product-analytics - Product analytics events (send, query scalar/timeseries)
  • app-builder - Low-code app management (list, get, create, update, delete, publish, unpublish)

Global Flags

Available on all commands:

--config string      Config file path (default: ~/.config/pup/config.yaml)
--site string        Datadog site (default: datadoghq.com)
--output string      Output format: json, table, yaml, csv, tsv (default: json)
--jq string          Filter/transform output with a jq expression (applied before formatting)
--verbose            Enable verbose logging
--yes                Skip confirmation prompts
--read-only          Block all write operations (create, update, delete)

--jq filtering

--jq applies a jq expression to the raw JSON response before output formatting, so it works with every -o format:

# Extract a single field across all monitors
pup monitors list --jq '.[].name'

# Select matching records and then format as a table
pup monitors list --jq '.[] | select(.name | endswith("prod"))' -o table

# Compose with other jq features
pup logs search --query="status:error" --jq '.data | length'

Cardinality: the jq expression may produce a stream of values.

  • 0 outputs → null
  • 1 output → the value (unwrapped)
  • 2+ outputs → an array

Filter target: --jq runs on the raw API response body (what --no-envelope prints). In JSON and YAML output its result becomes the envelope's data, and metadata (including truncated and next_cursor) is kept, so a filter never hides that more results exist. In table, CSV, and TSV output the jq result is rendered as rows as-is, and the metadata goes to stderr. pup agent schema <command> reports the starting expression for each command as returns.jq_root.

Limitation: commands that print output directly (e.g. pup auth login, some runbook steps) bypass format_and_print and do not honor --jq.

Recent Enhancements

Notebooks — Markdown representation (experimental)

notebooks get, create, update, and edit accept --markdown to work with a notebook as a Markdown document instead of a JSON cells array. These call /api/unstable/notebooks, which the notebooks team has not yet promoted to /api/v2; the path and response contract may still change.

  • get --markdown — print the notebook as Markdown (YAML frontmatter + body)
  • create --markdown --file doc.md — create from a Markdown file; prints the new id. Optional --name sets the notebook name (YAML frontmatter title, the Markdown form of data.attributes.name). Without --name, the first ATX heading (#–######, any level; setext headings are ignored) is used. If there is no heading either, the name is left empty.
  • update --markdown --file doc.md — replace the whole document
  • edit --markdown --file fragment.md — append the fragment server-side

create --markdown prints the id rather than the document, so it composes:

ID=$(pup notebooks create --markdown --file doc.md)
# or: pup notebooks create --markdown --file doc.md --name "Weekly review"
pup notebooks get "$ID" --markdown

The id is only available from the JSON:API representation — it is resource identity and deliberately absent from Markdown frontmatter — so create negotiates JSON:API and reports the id, leaving the document to get.

Constraints worth knowing before relying on these:

  • Rich-text notebooks only. Notebooks created through the older cells API have no Markdown projection, and the API returns an error for them.
  • update --markdown is lossy. It replaces the entire document, and anything Markdown cannot represent is dropped. The JSON path preserves more.
  • No conflict detection. document_revision is returned but not enforced by the API, so concurrent writers can overwrite each other on any write path.
  • No targeted edits. update replaces and edit appends; there is no way to modify one section in place.
  • --jq is rejected with --markdown, since the output is not JSON.
  • --output has no effect under --markdown: the document is printed as-is, with no format conversion.

v1.13.x — Session Replay API Support (#182)

  • rum replay segments get — fetch replay recording segments for a session view (--session-id, --view-id, optional paging/source flags)
  • rum playlists — extended with create, update, delete, and sessions list|add|remove|bulk-remove
  • rum viewership — new subgroup: history list, watch create|delete, watchers list

v0.64.x — Error Tracking Issue Filters (SDK PRs #1568, #1480)

  • error-tracking issues search — new optional filter flags:
    • --state <STATE> — filter by issue state: OPEN, ACKNOWLEDGED, RESOLVED, IGNORED, EXCLUDED
    • --team <UUID> — filter by team UUID assignee
    • --assignee <UUID> — filter by user UUID assignee
    • These flags are independent of the existing --track/--persona mutual exclusion

v0.34.1 — ACP Server (Datadog AI Agent Integration)

  • ✅ acp (new) — Local ACP + OpenAI-compatible server that proxies to Datadog Bits AI
    • serve — Start the server (default port 9099)
    • serve --agent-id <uuid> — Target a specific Datadog Bits AI agent (auto-discovers if omitted)
    • serve --port 8080 --host 0.0.0.0 — Custom bind address
    • Implements Agent Communication Protocol (ACP) at POST /runs and POST /runs/stream
    • Also exposes OpenAI-compatible POST /chat/completions and GET /models for tools like opencode
    • Requires OAuth2 (pup auth login) with notebooks_read + notebooks_write scopes

v0.33.4 — IDP Commands for Service Catalog

  • ✅ idp (new) — Agent-native access to the Datadog Service Catalog
    • assist <entity> — full context: owner, on-call, health, dependencies, metadata gaps, and suggested next actions
    • find <query> — search entities by name (defaults to kind:service)
    • owner <entity> — ownership + on-call responders for an entity
    • deps <entity> — upstream/downstream service dependencies
    • register <file> — POST v1–v3 YAML/JSON definitions to the Catalog entity API; YAML may contain multiple documents

v0.28.0 — New Command Groups and Full Pipeline Implementation

  • ✅ llm-obs (new) — LLM Observability: projects (create, list), experiments (create, list, update, delete, summary, events (list, get, submit), metric-values, dimension-values), datasets (create, list, batch-update, clone, restore, records, records-all, records-full), spans (search)
  • ✅ reference-tables (new) — Reference table management (list, get, create, batch-query)
  • ✅ obs-pipelines (upgraded from placeholder) — Full CRUD: list, get, create, update, delete, validate
  • costs — Added cloud cost configs: aws-config, azure-config, gcp-config (list, get, create, delete each)

v0.27.0 — Major Expansion

  • ✅ status-pages (new) — Status page management (pages, components, degradations CRUD)
  • ✅ code-coverage (new) — Code coverage summaries (branch-level and commit-level)
  • ✅ hamr (new) — High Availability Multi-Region connections
  • integrations — Added Jira integration (accounts, templates CRUD) and ServiceNow integration (instances, templates, users, assignment groups, business services)
  • cloud — Added OCI integration (tenancy configs CRUD, products)
  • synthetics — Added suites management (V2 API: search, get, create, update, delete)
  • synthetics — Added downtime management (V2 API: list, create, delete)
  • security — Added content packs (list, activate, deactivate), bulk rule export, and entity risk scores
  • incidents — Added global settings, handles, and postmortem template management
  • cases — Added Jira/ServiceNow issue linking, case project moves, and notification rules
  • cicd — Added DORA deployment patching and flaky tests management
  • slos — Added SLO status query (V2 API)
  • rum — Replaced playlist/heatmap placeholders with working RUM Replay API implementations

Widget snapshots

Render a bare widget definition with POST /api/v2/snapshot. The API queues rendering and returns a stored-image URL immediately. OAuth2 and API keys are supported; the organization must have the graphing-snapshot-api feature enabled.

# widget.json contains the definition, not the dashboard widget wrapper:
# {"type":"timeseries","requests":[{"q":"avg:system.cpu.user{*}"}]}
pup snapshots create --file widget.json --from 1h

# Wait for the PNG and save it as an Obsidian attachment:
pup snapshots create --file widget.json --from 1h --out cpu.png

# Return a URL that can load without Datadog authentication:
pup snapshots create --file widget.json --public --ttl 30d

# Accept a definition on stdin:
cat widget.json | pup snapshots create --file - --out cpu.png

URLs require Datadog authentication by default. --public makes the image accessible to anyone with its URL. Use a local PNG for an Obsidian attachment that should survive the server-side retention period. The image is a static snapshot; embedding its URL does not refresh its data.

Without --out, the returned URL may still be pending. With --out, Pup polls every two seconds until it receives PNG bytes, with a default timeout of 120 seconds (--timeout). Errors and timeouts retain the URL so you can retrieve the same snapshot later. Existing output files are never overwritten.

Defaults are --from 1h --to now --width 1000 --height 400 --ttl 60d. Times accept relative values, RFC3339, or Unix seconds/milliseconds. Width must be 32–2000 pixels and height 32–1000. Retention accepts 30d, 60d, 90d, 1y, 2y, or inf.

Use --additional-config context.json for a JSON object containing API context, such as notebook_id for notebook-local datasets or template_variables:

{
  "notebook_id": 123,
  "template_variables": [
    {"name": "env", "prefix": "env", "values": ["prod"]}
  ]
}

The backend validates supported widget types and data access. This command creates a stored artifact and is blocked by --read-only.