dt-sdlc-observability-design

Generates observability plans with baselines, blast-radius maps, verification DQL, and rollback criteria from Dynatrace data.

Updated Dec 12, 2025
One-click install
npx skills add https://github.com/ricardojjulia/ESACompanion --skill dt-sdlc-observability-design-ricardojjulia
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: dt-sdlc-observability-design
Source: https://github.com/ricardojjulia/ESACompanion/tree/main/.github/skills/dt-sdlc-observability-design
Command: npx skills add https://github.com/ricardojjulia/ESACompanion --skill dt-sdlc-observability-design-ricardojjulia

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Teams often add monitoring after code is written, leaving them without baselines, verification queries, or measurable rollback criteria when deploying changes. This Skill designs observability before implementation by grounding every decision in live Dynatrace runtime data. ## Core Features & Use Cases - Baseline Discovery: Runs DQL queries against live data to capture latency percentiles, error rates, throughput, and infrastructure saturation for a target service. - Blast Radius Mapping: Uses topology and trace data to identify upstream consumers, downstream dependencies, shared infrastructure, and active problems a change could affect. - Verification & Rollback Design: Produces ready-to-run post-deployment DQL queries, SLO recommendations, and rollback thresholds derived from actual baseline values. - Use Case: Before modifying a checkout service, a team uses Dynatrace Assist to resolve the service entity, collect its p90 latency and error-rate baseline, map its five upstream consumers, and export a Markdown observability plan with regression checks and rollback triggers for the development team. ## Quick Start Ask Dynatrace Assist to build an observability plan for the service you are planning to change, starting with baseline collection and blast-radius mapping.

Frequently Asked Questions about dt-sdlc-observability-design

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create an observability plan before deploying a service change?▼

Start in Dynatrace Assist by naming the service you plan to change. The skill resolves the entity, collects baseline latency and error metrics via DQL, maps upstream and downstream dependencies, then generates verification queries, SLO recommendations, and rollback thresholds in an exportable Markdown plan.

How to map blast radius of a service change with Dynatrace?▼

Blast radius mapping queries span data to find upstream callers (client spans with the target as peer.service), downstream dependencies, and co-located services on shared hosts. It also checks for active Davis problems on affected entities before the change proceeds.

Does this skill require source code access?▼

No. The primary context is Dynatrace Assist in the platform UI, where all analysis runs against live runtime data already collected by Dynatrace. An optional IDE/agent mode with source access adds code-level instrumentation review and CI/CD integration.

What DQL queries are used for baseline collection?▼

Baseline queries use the timeseries command with dt.service.request.response_time percentiles (p50, p90, p99), request and failure counts for error rate, plus host CPU and memory usage metrics. Results are documented in a baseline table for later regression comparison.

When should I not use observability-driven design?▼

If you only need an automated post-deploy check, start with a quality-gates or post-deployment validation skill instead. This skill is intended for teams ready to design observability before implementation, not for simple deployment verification.

Why must entity IDs be resolved before running queries?▼

DQL queries filter on Smartscape entity IDs, so placeholder values like <service-id> will fail or return nothing. The skill resolves service names to real entity IDs first using smartscapeNodes queries, then reuses those IDs in all subsequent analysis.