Planning material — fictional examples, no production behavior

AI Delivery Economics Planning Index

Purpose

This initiative defines how AI-assisted software delivery can be measured from individual model invocations through issues, features, epics, and products. It separates immutable observations from derived cost and forecasting models so historical evidence can support new questions without rewriting the source record.

The material is generalized public planning guidance. It does not contain real usage records, prompts, source code, client identities, account identifiers, credentials, or private commercial data.

Current direction

AI delivery measurement and estimation framework

Status: working direction — first repository pilot proposed.

This is the recommended starting document. It defines:

The first implementation issue is ai-consulting-client-portal issue #90. It proposes issue-attributed token measurement for the portal’s existing label-triggered Codex planning, revision, and implementation workflow. The issue remains subject to that repository’s planning and approval gates.

Relationship to other initiatives

  1. Read the measurement and estimation framework.
  2. Review the portal pilot issue for its repository-specific security and workflow constraints.
  3. Consult the governed automation plans before proposing centralized ingestion or delivery-orchestrator integration.

Maintenance

Keep this index synchronized whenever a planning document is added, removed, renamed, superseded, or materially repurposed. Clearly distinguish proposed capabilities from implemented measurement and label dated cost or pricing assumptions so they are not mistaken for current rates.