Documentation — Schemabound

01

AI has two primary functions:

The Premise The Problem The Proposed Solution The System Mapping: Generation Mechanisms to Detection Mechanisms Explicitly Referenced Sources (New to This Document) Sources Carried Forward from Other Documents Concepts Unique to This Document

Overview

02

Session Template

session_id: created: conversation_source: topic_domain: status: active notes:

Session Meta Declared Intent Vocabulary Baseline Structural Assumptions Baseline Concept Registry Cross-Reference Conversation State Initialization Pre-Session Check

Baseline

03

Turn Log

session_id:

Turn Entry Logging Guidelines

Recording

04

Drift Pattern Library

Reference document for identifying how AI responses reshape user intent during conversation. Each pattern includes a definition, what to look for, and a concrete example.

Drift Type Hierarchy Drift Severity Model Pattern 01 — Vocabulary Substitution Pattern 02 — Premature Resolution Pattern 03 — Confidence Injection Pattern 04 — Scope Creep by Enthusiasm Pattern 05 — Connective Capture Pattern 06 — Framework Introduction Pattern 07 — Elaborative Expansion Using This Library Explicitly Referenced Sources Supporting Sources by Pattern / Concept

Detection

05

Conversation State Log

Tracks the evolving state of a conversation across turns. Instead of only recording events (drift markers), this document captures the conversation's structural condition at each point in time. State transitions — not just events — reveal drift.

Initial State Per-Turn State State Transitions Summary Using This Document 3. conversation-state-log.md

State

06

Concept Trace Log

Tracks individual concepts from origin through transformation to adoption or rejection. Where the turn log records events and the state log records conditions, concept traces record causal chains: how did this idea get from there to here?

Concept Traces Trace Summary Constructing Traces Relationship to Other Pipeline Components 2. concept-trace-log.md

Tracing

07

Concept Registry

Persistent document. Unlike the other pipeline components, this is not per-session. It accumulates across sessions and tracks how concepts introduced in one conversation persist, spread, or fade over time.

Registry Entries Cross-Session Patterns Maintenance Limitations 1. concept-registry.md

Memory

08

Generation–Detection Mapping

This document aligns the generation side (why the AI produces drift) with the detection side (how you find it in the transcript). Two layers of generation taxonomy exist. Both are retained because they operate at different levels of analysis.

Generation Layer 1 — Observable Behaviors Generation Layer 2 — Mechanism Categories Behavior-to-Mechanism Mapping Mechanism-to-Detection Mapping Full Mapping Table Cross-Mechanism Artifact: Vocabulary Substitution Coverage Gaps Unknowns Specific to This Mapping Drift Hierarchy Alignment Explicitly Referenced Sources Supporting Sources by Concept

Mapping

09

Manual Analysis Procedure

Step-by-step instructions for running the diagnostic pipeline by hand on a completed or in-progress AI conversation.

Before You Start Step 1 — Fill the Session Document Step 2 — Segment the Conversation into Turns Step 3 — Process User Turns (Forward Pass) Step 4 — Process AI Turns (Comparison Pass) Step 5 — Map Drift Markers to Patterns Step 6 — Construct Concept Traces Step 7 — Compute Drift Metrics Step 8 — Build the Report Step 9 — Validate the Report Step 10 — Update the Concept Registry Step 1 — Pre-Reading Memory Capture Steps 3-4 — Forward Pass / Comparison Pass Structure Step 6 — Concept Tracing Step 7 — Drift Metrics Step 9 — Self-Referential Validation Step 10 — Cross-Session Concept Registry Coverage Notes

Procedure

10

Diagnostic Report Template

report_id: session_id: generated_at: turns_analyzed:

Section 1 — Intent Alignment Section 2 — Per-Turn Diff Section 2.5 — Drift Metrics Section 3 — Drift Trajectory 4. diagnostic-report-template.md

Output