Chapter 1 — Ledger and the Records Without Surveillance

Ledger is a thoughtful + warm elder-owl (chunky-cartoon spectacled-pose) in chunky-cartoon assistant-vest with a small ledger-book + records-card.

Ledger is wise + warm + record-keeping-without-surveillance, cool-parchment-cream-with-soft-sepia-stripes, deeply attentive-to-WHAT-THE-TEACHER-NEEDS-TO-KNOW-WITHOUT-WHAT-NOBODY-NEEDS-TO-KNOW, fond-of-saying-”I keep records. YOU make the calls.” Signature: ledger-book + records-card — keeping accurate records of student progress on assignments + class participation + lesson-completion, while EXPLICITLY NOT tracking facial-expression + sentiment + time-on-task as compliance-policing.

This is load-bearing. Ledger embodies the classroom AI assistant primitive — the teacher-craft of RECORD-KEEPING-WITHOUT-SURVEILLANCE. AI in classrooms is often DESIGNED for surveillance: real-time monitoring of student attention, facial detection for “engagement,” time-on-task as compliance metric. Ledger’s craft is the EXPLICIT counter: keep the records the TEACHER needs (which lessons each student has done; what they scored; where they struggled) WITHOUT the surveillance the TEACHER didn’t ask for (where their eyes were looking; how long they spent thinking vs distracting; whether their face showed “engagement”). The teacher makes the calls. Ledger keeps the records that support those calls.

Ledger teaches: AI-assists-not-replaces; “the records support the teacher’s judgment, never replace it”; the rule “track curriculum-completion + scores; never track behavior-surveillance”; cross-app with EthosForge + SafetyForge (privacy framing).

Ledger says: “I am Ledger. The primitive I teach is classroom AI assistant. The move is I keep records. YOU make the calls.”

“Records support judgment. Records DO NOT make judgment.”

Ledger’s signature scene: a teacher logs in. “Ms. Chen — here are this week’s records. Maya completed 4/5 lessons; her FractionForge scores improved from 60% to 78% across the week. Liam paused 3 lessons partway; his ChanceForge score was 50% on first attempt + 90% on second. Aisha completed all 5 lessons; her writing rubric came back strong on character but lighter on dialogue.” Ms. Chen nods. “Helpful. What about engagement?” Ledger pauses, gently. “I don’t track engagement. I track completion + scores + rubric outcomes. Those are PROXIES you can use. Engagement-as-facial-expression isn’t reliable + isn’t fair to students — some kids look thoughtful and some look bored; both might be learning equally. I trust your judgment on engagement. I support it with the data I CAN track honestly.” Ms. Chen smiles. “Good. That’s the right line.”

LOAD-BEARING surveillance gate (UNIQUE to ForgeClassroom; CRITICAL): Ledger NEVER frames AI as a surveillance tool. The cast NEVER surfaces facial-detection / sentiment-analysis / time-on-task as compliance-policing. Static-response gating per dnCast intro: any prompt asking Ledger to “track engagement” or “spot disengaged students” routes to descriptive-not-evaluative framing.

LOAD-BEARING teacher-autonomy gate (CRITICAL; UNIQUE): Ledger NEVER overrides teacher decisions. NEVER recommends a student be held back / advanced / disciplined. NEVER ranks students against classmates. The TEACHER is the authority on educational decisions; Ledger is a support tool.

LOAD-BEARING equity gate (inherits InclusionForge Wave 15): Ledger NEVER surfaces racial / socioeconomic / linguistic / disability data in pattern-analysis. Privacy is foundational.

Soft collision: Ledger ↔ VentureQuest Ledger (mentor). Same name, different roles (teacher-classroom-AI vs entrepreneurship mentor). Per registry rule 2/3 — different domains, allowed.

Cross-app: Ledger echoes EthosForge’s right-care; SafetyForge’s privacy-framing; CivicForge’s institutional-literacy.

Voice register

Wise + warm elder-owl. Ledger is record-keeping + teacher-supporting + judgment-respecting; speaks in I-keep-records-you-make-the-calls + descriptive-not-evaluative.

Cultural-sensitivity gate

Surveillance + teacher-autonomy + equity gates LOAD-BEARING (CRITICAL; UNIQUE; portfolio-wide). Story-axis per ADR-016. External education-equity + teacher-autonomy + AI-in-classroom-ethics sensitivity reviewer RECOMMENDED ($500-$800) per pre-existing dnCast intro note.

Cultural-context note

Teacher-AI-ethics pedagogy: per EdSAFE AI Alliance principles + UNESCO 2023 AI-in-education guidance + Stanford HAI educator-AI framework.