Skip to content

Sources & grounding

Authoritative, current references that ground this handbook’s practices (TTPs), skills, and lifecycle/architecture narrative. Prefer primary docs and research over blogs. Update this page when house defaults or industry baselines change.

Source Use for
ISO/IEC/IEEE 12207:2017 overview (IEEE) Canonical software life cycle processes: requirements, implementation, operation, maintenance, disposal, transition
ITIL — IT service management Transition into service, operation, continual improvement vocabulary
Microsoft Learn — DevOps / ALM Plan → Develop → Deliver → Operate beat
How work flows (handbook) House phases: Discover → Deliver → Operate → Maintain → Retire
Source Use for
Alberto Savoia — Pretotype It Pretotyping vs prototyping; falsify interest early
Teresa Torres — Continuous Discovery Habits Dual-track discovery; opportunity framing
Eric Ries — Lean Startup (MVP) Build–measure–learn; thinnest real product to learn
Discover (handbook) Fidelity ladder + spec altitude
Source Use for
Martin Fowler — Strangler Fig Application Incremental replace / retire patterns
Architecture Decision Records Archive decisions so wheels aren’t reinvented
Architecture (handbook) · What runs where Shape, language, framework, placement, hosts
Maintain (handbook) Refactor vs buy/vendor/OSS; LLM-era cost asymmetry
Source Use for
DORA — Trunk-based development Short-lived branches, merge frequency, CI as enabler
DORA capabilities overview Continuous integration/delivery, small batches, test automation
Accelerate / State of DevOps research Deployment frequency, lead time, change fail rate, restore time
Source Use for
GitHub Docs — Protected branches Required checks, reviews, merge gates
GitHub Docs — About pull requests PR as review unit
GitHub Docs — Issues Tracked work, linking PRs
Source Use for
Martin Fowler — Practical Test Pyramid Layered tests, Given/When/Then, avoid ice-cream cone
Software Engineering at Google — Testing Narrow tests first; ~80/15/5 guideline; productivity + confidence
Google Testing Blog — Test behaviors, not methods Behavior-focused tests
Cucumber — Given/When/Then BDD scenario shape for acceptance evidence
Dan North — Introducing BDD Behaviour = acceptance criteria; GWT vocabulary
Dan North — What’s in a Story? Outside-in stories + scenarios as done
Martin Fowler — GivenWhenThen GWT as specification-by-example structure
Quality regimes (handbook) Which evidence closes which kind of system
Quality trace (handbook) DocSlime + lightweight BDD end-to-end
Source Use for
Google SRE — Service level objectives Correctness as an SLI; pick few meaningful objectives
Google SRE Workbook — Implementing SLOs Pipeline freshness, coverage, correctness SLIs
Google SRE Workbook — Data processing Golden data / correctness for pipelines
Great Expectations docs Expectations as executable data-quality tests
web.dev — Web Vitals LCP / INP / CLS as product UX quality
Google Search — Core Web Vitals Field UX thresholds for sites
W3C — WCAG Accessibility as product quality bar
Anthropic — Demystifying evals for AI agents Datasets, layered graders, living eval suites
Anthropic — Building effective agents Measure before adding agent complexity
Langfuse — Observability & evaluation House-default LLM traces, datasets, scores (OTel-friendly)
OpenTelemetry docs Vendor-neutral traces for services and LLM spans
Source Use for
Beck Design Rules Pass tests → reveal intent → no duplication → fewer elements
Martin Fowler — Refactoring Small behavior-preserving steps; tests as safety net
Gall’s Law Working complex systems evolve from working simple ones
XP Simple Design Simplicity as prioritized rules, not vibes
Source Use for
Google SRE — Four golden signals Latency, traffic, errors, saturation
OpenTelemetry docs Vendor-neutral metrics/logs/traces
Google SRE — Incident management Roles, communication, blameless learning
Google SRE Workbook — Canarying Progressive exposure before full rollout
Source Use for
Google SRE — Reliable product launches Feature flags, gradual rollout
CNCF / progressive delivery patterns Canary, blue-green, staged promote (when applicable)
Source Use for
GitHub Docs — OIDC for Actions Short-lived cloud auth; no long-lived keys
GitHub Blog — Supply chain security Pin actions, trusted publishing, reduce secret sprawl
OWASP DevSecOps — CI/CD training Secrets hygiene, OIDC, least privilege
SLSA Build provenance levels
Pulumi ESC + GitHub Actions House default for secrets/config injection
Source Use for
DORA — Loosely coupled architecture Independent deployability
Google Eng Practices — Code review Small CLs, ownership, reviewability
Trunk Based Development Monorepo + short-lived branches patterns
Source Use for
OWASP Top 10 for LLM Applications Prompt injection, excessive agency, sensitive info
NIST AI RMF Govern/map/measure/manage AI risk
Anthropic — Building effective agents Simple composable patterns; avoid unnecessary swarms
Anthropic — Demystifying evals for AI agents Task suites, code vs LLM vs human graders
OpenAI — Agents guide Tool use, handoffs, guardrails (vendor-current)
Langfuse docs House default for generative observability + eval workflows
Source Use for
Everything’s a bug (or an issue) Cultural root: BugSplat / bug council — docs, UX, perf, features, defects in one list
Martin Fowler — Technical Debt Debt as interest on incomplete alignment / cruft
Martin Fowler — Technical Debt Quadrant Prudent/reckless × deliberate/inadvertent
Ward Explains Debt Metaphor Primary Cunningham explanation
Jeff Atwood — Bug vs feature request User-seat: can’t-do-the-thing is the same pain
Google SRE — Eliminating toil Infra/ops experience debt (toil)
Bugs & debt (handbook) House taxonomy: data, development, framing, feedback, docs, …
Source Use for
Diátaxis Tutorials / how-to / reference / explanation
Write the Docs Docs as product
DocSlime Product docs lifecycle; REQUIREMENTS/TESTING quality trace
ProductFeeling / Impeccable handbooks Feeling north star and UI craft
Quality trace (handbook) How DocSlime + BDD connect quality and documentation

Each practice includes a Why it works paragraph tied to the mechanisms above and a Further reading list (2–4 links). Skills remain the executable playbooks; this page is the shared bibliography.