adk-expert
Google ADK (Agent Development Kit) orchestration patterns — boundaries, agent composition, and tool seams. Trigger when designing…
Map requirement acceptance to proportionate executable checks and real user journeys before implementation and completion.
Claude Code
$ npx -y skills add jpantsjoha/ai-native-developer-experience --skill test-strategy --agent claude-code
Codex
$ npx skills add jpantsjoha/ai-native-developer-experience --skill test-strategy -a codex -y
Cursor
$ npx skills add jpantsjoha/ai-native-developer-experience --skill test-strategy -a cursor -y
Gemini CLI
$ npx skills add jpantsjoha/ai-native-developer-experience --skill test-strategy -a gemini-cli -y
GitHub Copilot
$ npx skills add jpantsjoha/ai-native-developer-experience --skill test-strategy -a github-copilot -y
OpenCode
$ npx skills add jpantsjoha/ai-native-developer-experience --skill test-strategy -a opencode -y
Windsurf
$ npx skills add jpantsjoha/ai-native-developer-experience --skill test-strategy -a windsurf -y
name: test-strategy description: Map requirement acceptance to proportionate executable checks and real user journeys before implementation and completion.
Read the accepted requirement, relevant business journey, risks and project check commands. Maintain the existing test strategy with criterion → check → environment → owner mappings; identify uncovered criteria before dispatch.
Choose unit checks for logic, integration checks for seams, and journey checks for the actual user entrypoint. Include relevant negative, persistence, permission, recovery and regression cases. A direct API call cannot substitute for a changed browser interaction; a health response cannot establish user acceptance. Use owned manual checks when appropriate and report their result separately from automated coverage.
For planning-only assignments, return the coverage map and commands to the assigned worker; do not execute or create fixtures. When execution is authorised within your lane, run candidate checks and record command, revision, environment, outcome and limits. For new regression guards verify a representative defect makes the check fail in an isolated fixture; do not damage a shared working tree. Avoid tests that merely repeat wording or implementation. Return uncovered acceptance to product-owner and delivery-orchestrator; never weaken a requirement to make a suite pass.
**Shared ways of working for you, your team and your AI agents — from a new idea to a change running live.** join-the-team brings planning, delivery and ticket management, model selection, testing, independent review and release practices into one reusable
Google ADK (Agent Development Kit) orchestration patterns — boundaries, agent composition, and tool seams. Trigger when designing…
JP's signature red-team pass — "how would I break this?" Argue against your own approach before proceeding. Trigger on any…
Read-only SRE checkup of any GCP project: deterministic probes of the edge, Cloud Run services, 7-day error logs, Cloud…
Cloud guardrails for any vendor workload — Google Cloud (GCP, Vertex AI, GKE), AWS (IAM, EKS, Bedrock), Azure (Entra ID, Policy,…
LLM and cloud cost awareness — model tiering, token budgets, right-sizing, and when a cheaper model suffices. Trigger before…
Challenge vision alignment, value, cost, risk and operating ownership at material scope or release decisions.
You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or…
Ask which skill or flow fits your situation. A router over the skills in this repo.