II.
Workflow overview
Reference · liveworkflow:prompt-regression-testing
Prompt Regression Testing overview
Runs automated evaluation suites against LLM prompts and chains on every change — comparing output quality, latency, cost, and safety metrics against baseline snapshots and flagging regressions. Excludes prompt authoring.
Attributes
displayName
Prompt Regression Testing
workflowKind
development
triggerType
event-driven
typicalCadence
per-pull-request
complexity
single-team
description
Runs automated evaluation suites against LLM prompts and chains on every
change — comparing output quality, latency, cost, and safety metrics
against baseline snapshots and flagging regressions. Excludes prompt authoring.
Outgoing edges
applies_to_domain2
- domain:ml-ops·DomainMLOps
- domain:software-engineering·DomainSoftware Engineering
involves_role2
- role:ml-engineer·RoleMachine Learning Engineer
- role:ai-champion·RoleAI Champion
performed_by_org_unit2
- org-unit:ai-enablement·OrgUnitAI Enablement
- org-unit:ml-team·OrgUnitML Team
requires_skill_area3
- skill-area:prompt-engineering·SkillAreaPrompt Engineering
- skill-area:eval-driven-development·SkillAreaEval-Driven LLM Development
- skill-area:agentic-loops·SkillAreaAgentic Loops
triggers_responsibility2
- responsibility:ai-safety-guardrails·Responsibility
- responsibility:ai-agent-usage-review·Responsibility
Incoming edges
None.