CALIBER
Quickstart
Use CALIBER

Prompts

The task-oriented entry point for prompt assets in CALIBER: author, evaluate, calibrate, release, and recover without starting from deep architecture.

System userDeveloperEvaluatorHow ToGA
PrerequisitesA CALIBER deployment with prompt access
Reviewed 2026-08-10 · current main branch docs contract

This page is the task-oriented entry point for using prompt assets in CALIBER. Use it before the deep prompt architecture page when the question is practical: author, evaluate, calibrate, release, or recover.

At a glance

TaskStart hereDeep reference
Edit a prompt safelycreate or update a prompt revisionPrompts architecture
Measure prompt qualityrun evaluation against a test setEvaluation
Generate better candidateslaunch prompt calibrationCalibration
Make a candidate liveuse the explicit release/apply pathThe refinement loop
Recover from release troubleuse the runbookOperations runbook

1. What prompts are for in CALIBER

Prompts are governed assets with version history, evaluation context, and an explicit release boundary. Editing a prompt does not make it live by itself.

That distinction matters because CALIBER separates:

  • authoring
  • evidence generation
  • candidate generation
  • live release

2. Common tasks

You want to...Read this next
create or revise a promptPrompts architecture
compare prompt behavior against known examplesEvaluation and test sets
generate improved candidatesCalibration
understand release, rollback, and review boundariesReview and release flows

3. What usually confuses readers

SymptomWhat it usually means
The prompt changed, but production behavior did notthe change is not live until the explicit release path completes
A candidate scored well, but nothing deployedevaluation or calibration produced evidence, not live apply
Prompt behavior regressed after a releaseinspect the release path, evidence set, and rollback options

CALIBER : Contextual Adaptive Lifecycle for Intelligent Build, Evaluation, and Refinement — this page is generated from the authoritative Markdown sources in docs/ and the repository-level ARCHITECTURE.md.