Guides · APRF practice notes
AI Lifecycle Guides
Treat prompts and context as production artifacts. These guides support APRF Prompt Engineering, Context Engineering, Model Governance, and Evaluation.
Prompts, context, models, and evals — APRF Model & Prompt Lifecycle.
For RAG corpus and privacy controls, see Data guides.
AI Production Readiness Framework
Related APRF controls
- AI Evaluation Gates Before Production Release
If you can change a prompt or model without a failing test suite blocking the deploy, you do not have an evaluation gate. APRF expects measurable pass conditions before production.
- Context Packing, Truncation, and Trust Boundaries
Everything in the context window can steer the model. APRF Context Engineering requires trust labels, packing rules, and truncation that does not drop safety or authorization critical text.
- Model Versioning and Rollback in Production
"Latest" is not a production version. Pin model IDs, record what served each request, and practice rollback like any other dependency.
- Version Prompt and Context Artifacts in Production
If prompts live only in a Slack paste or an untracked YAML on one laptop, you cannot roll back a bad change. Production AI needs prompts and context packs as versioned artifacts.
Next: Prompt Engineering
Open the related pillar specification for mandatory checks, artifacts, and pass conditions. Self-attest is optional.