cc-058 · GAThe academy track of BottegaTrains Agent Operator Experts (AOEs)SHA-256 provenance

Agent Operator Academy

The academy track of Bottega — Claresia's master-apprentice studio. Hands-on curriculum that moves workers from todayAI-adoptedAgent Operator Expert (AOE). Cross-functional capability via the 5 GA atlases (cc-058) + vertical depth via 8 Bottega specialty courses (cc-084). Same OB 3.0 contract, two issuer DIDs, one verifier.

5
Capability atlases
AE · PM · MKTG · FP&A · Eng Lead
8
Bottega specialties
Sales · Finance · Marketing · Ops · …
2
Delivery modes
Embedded · Standalone

Five core tenets

The non-negotiables that separate Agent Operator Academy from a classic LMS.

Hands-on only
No multiple-choice. Every learning event produces a real artifact, prompt, workflow, or KPI movement scored against rubrics derived from cc-052 hero-scene patterns.
Embedded in the daily LLM
The journey runs inside whichever LLM the worker already uses in production — Claude (native), M365 Copilot, ChatGPT Enterprise, Claude Enterprise. Never a separate web app.
Multi-dimensional skill mapping
Curriculum is generated per (function × role × business_model × industry × company_type) from the cc-050 / cc-051 knowledge graph. No two workers see the same journey.
Two delivery modes
Embedded for Claresia customers (production telemetry attestation). Standalone for non-customers + individuals (synthetic-data practice). Same canonical journey, same final credential.
Forge Academy DNA + LMS spine
Reuse Forge's exercise/coaching engine (fg-002 interactive artifacts, fg-005 AI Coaching, fg-009 progress tracking). Layer LMS-grade attestation, audit log, and verifiable credentialing on top.

Seven product pillars

Skill DAG → embedded runtime → practice library → attestation → standalone mode → credentialing → operator console.

1
Skill-Mapping Engine
Per-archetype skill DAG with prerequisites, evidence requirements, and proficiency thresholds. Addressable by the cc-050 5-tuple atlas key.
2
Embedded Journey Runtime
Adapter layer surfaces the personalised journey inside Claude, Copilot, ChatGPT, Claude Enterprise. Same canonical state machine, multiple host renderings.
3
Hands-On Practice Library
Tasks generated from cc-051 archetype runtime spec — real cadences, real artifact templates, real KPIs. Worker performs the task on real or synthetic data.
4
Production-Mode Attestation
In-production telemetry hook (cc-050 Hub adapter) reads the worker's real symbolic event stream and matches it against the skill DAG. No separate test required.
5
Standalone Mode
Public enrolment portal. Same journey runtime, but the worker connects their own LLM and practices on synthetic data. Same final certification.
6
Credentialing & Audit
Per-level credentials (BASELINE / AI_ADOPTED / AGENT_OPERATOR) issued on attestation. Open Badges 3.0, JSON-LD, LinkedIn-importable, employer-verifiable.
7
Academy Operator Console
Customer-side dashboard extending cc-053 Maturity Engine — per-role progress, bottleneck identification, cohort analytics, intervention recommendations.

Dependency graph

cc-058 sits on top of the Synthetic Twin Trilogy + Forge platform — no forks, all reuse.

cc-050
Intelligence Hub
canonical schema · provenance contract · typed Hub records
cc-051
Universal Workers Behavior Layer
archetype runtime spec · cadence + artifact templates
cc-052
Synthetic Org Simulator
hero-scene rubric · symbolic event substrate
cc-053
Maturity Engine
L0 / L1 / L2 anchors · per-role transforms
fg-002
Interactive Artifacts
exercise runtime — wrapped, not forked
fg-005
AI Coaching Layer
Socratic SSE coaching, drop-in
fg-009
Platform Foundation
JWT cross-domain auth · progress schema

GA progress · 8 milestones

M1–M7 shipped at GA. M8 Dainese pilot completed in standalone-mode (Dainese has not flipped cc-050 Option B yet — embedded-mode flip queued as cc-058 follow-up).

M1
M2
M3
M4
M5
M6
M7
M8
M1
Skill-mapping engine + canonical schema
5 atlases · cross-language hashes
M2
Hands-on practice library + RubricEngine
cc-052 hero-scene rubric reused
M3
Reference Claude embedded adapter
canonical state machine + Hub records
M4
Production-mode attestation pipeline
cc-050 Hub matcher · 6 typed kinds
M5
Standalone enrolment portal (Astro)
Vercel · synthetic-data lane
M6
Credentialing service (Open Badges 3.0)
JSON-LD · public verification endpoint
M7
Academy Operator Console (cc-053 extension)
cohort + bottleneck telemetry
M8
Dainese pilot — standalone mode against real org
5-incumbent cohort · 10 credentials issued
Provenance contract: SHA-256 over canonical JSON · identical to cc-050 / cc-051 / cc-052 / cc-053.
cc-058 dashboard · Electric Noir v2 · built with Astro + React + Recharts.