Appearance
Glossary
Quick definitions for terms used throughout this reference template. Record your org-specific definitions in your own glossary or ADRs.
| Term | Definition |
|---|---|
| ADR | Architecture Decision Record — documents a significant technical choice, status, and consequences |
| AIPO | AI Platform Owner — Bedrock/Q/MCP config, prompt policies, agent audit |
| ARB | Architecture Review Board — escalated cross-team design decisions |
| ABAC | Attribute-Based Access Control — permissions from attributes (tags, tenant, classification) |
| Accepted ADR | ADR status meaning the decision is approved and binding for implementation |
| Approved spec | Specification reviewed and signed off — required before SOP-004 |
| Canary deploy | Rolling out to a small prod slice before full promotion |
| CI / CD | Continuous Integration / Continuous Delivery (or Deployment) |
| Contract test | Test verifying API consumer/provider agreement (e.g. Pact) |
| DAST | Dynamic Application Security Testing — runtime security scanning |
| DATA (steward) | Role owning classification glossary and catalog quality |
| DEV | Developer — implements against approved spec |
| EM | Engineering Manager — escalation, capacity, non-security exceptions |
| Ephemeral env | Short-lived stack per PR for integration/E2E tests |
| Guardrail | Automated policy that blocks or warns on unsafe code or usage |
| Human-in-the-loop | Required human approval at defined gates — AI recommends, humans decide |
| IAM | AWS Identity and Access Management |
| Identity Center | AWS IAM Identity Center (SSO) for workforce access |
| IRSA | IAM Roles for Service Accounts — EKS pod-level AWS permissions |
| JIT | Just-in-time — temporary elevation of privileges |
| KMS | AWS Key Management Service — encryption key management |
| LF-tags | Lake Formation tags for fine-grained data lake access |
| MCP | Model Context Protocol — tool/context integration for AI agents |
| OIDC | OpenID Connect — federation for CI/CD to AWS without long-lived keys |
| OPA | Open Policy Agent — policy-as-code for IaC and pipelines |
| OTel | OpenTelemetry — vendor-neutral metrics, logs, traces |
| PO | Product Owner — priority and acceptance criteria |
| RACI | Responsible, Accountable, Consulted, Informed |
| RAG | Retrieval-Augmented Generation — LLM + document index |
| AI toolkit | Catalog of AI tools and adjacent artifacts (ADR, specs, MCP) |
| RBAC | Role-Based Access Control — permissions assigned to roles |
| RUM | Real User Monitoring — client-side performance and experience |
| SAST | Static Application Security Testing |
| SCA | Software Composition Analysis — dependency/CVE scanning |
| SEC | Security Engineer — threat models, scan policy, exceptions |
| Secrets Manager | AWS service for storing and rotating secrets |
| SLO | Service Level Objective — target reliability metric |
| SOP | Standard Operating Procedure — repeatable team process |
| Spec-driven | Implementation bound to an approved specification/contract |
| SRE | Site Reliability Engineering / platform — deploy, SLOs, incidents |
| SSM | AWS Systems Manager Parameter Store |
| SSO | Single Sign-On |
| STRIDE | Threat modeling framework (Spoofing, Tampering, …) |
| T1 / T2 / T3 | Service tiers — increasing rigor for deploy and on-call (GOVERNANCE) |
| TechDocs / Backstage | Developer portal patterns for catalog and docs |
See also
- GOVERNANCE — roles and decision rights
- Architecture — layer model
- Adoption roadmap