Competency Frameworks & Cross-Standard Equivalencies

Architecture: Open Assessment Standard (OAS v1beta1)

Purpose: This document explains how ColabEdu classifies every exercise and rubric along two independent dimensions — subject-specific skills and cross-subject transversal competencies — and how this dual classification enables content reuse across courses from different countries and programs.

---

Why Implement This System?

Before diving into the architecture, it is worth understanding what concrete problems this framework solves — and why the internal complexity is justified (since it is invisible to exercise authors and to the grading engine).

Authors do nothing extra

The fundamental design rule is: exercise authors never write competency tags. They only declare a task_type (e.g., comp.ap.slc.frq_task2) and the CompetencyClassifier automatically infers all tags — CEFR skills, Bloom levels, EU Key Competences — from rules defined in the taxonomy YAML files.

# What the author writes (a single field):
spec:
  exercises:
    - task_type: comp.ap.slc.frq_task2   # ← that's all

# What the system infers automatically:
# competency_tags:
#   skills: [presentational_writing, argumentation, source_analysis]
#   cefr:   [C1]
#   bloom:  [ANALYZE, EVALUATE]
# transversal_competency_tags:
#   eu_key_competences: [EU_CCL, EU_CPSAA]
#   lomloe_cc:          [CCL, CPSAA]

Standard-specific grading does not change

Competency tags are indexing metadata, not grading logic. The GradingOrchestrator always grades an AP exercise with an AP rubric and a LOMLOE exercise with a LOMLOE rubric. Tags do not alter any scoring criterion — they only enable the cross-standard use cases described below.

The six real benefits of the system

🌎 Cross-country content reuse

An AP FRQ Task 2 exercise tagged with argumentation / C1 can automatically cover the CE.LCL.B.1.4 gap (LOMLOE) in a Spanish curriculum bundle. No manual curation required.

📊 Student competency radar

The dashboard can display: “You are at B2 in reading comprehension but only B1 in oral production”. Actionable data for students and teachers.

🔍 Automatic curriculum gap analysis

The ACA automatically detects that an AP SLC bundle has 0 literary_analysis exercises and suggests importing IB Spanish A or LOMLOE C.1 specs.

🎯 Smart exercise recommendation

If a student fails argumentation in LOMLOE, the system can recommend AP FRQ Task 2 exercises — same competency, different context. Higher retention.

⚖️ AI grader calibration

The GradingOrchestrator knows that “C1 argumentation” requires a thesis, counterargument, and closing — regardless of whether the exercise is AP, IB, or LOMLOE.

📋 Compliance reporting

Automatic generation of EU Key Competences coverage reports for educational institutions. A B2B differentiator for administrations with European accreditation requirements.

Tip

Design principle: The system’s complexity lives inside the taxonomy YAML files and the CompetencyClassifier — not in exercise authors or grading rubrics. Rubrics remain 100% specific to each standard, country, and region.

---

The Core Insight: Two Orthogonal Dimensions

Every educational exercise works on two levels simultaneously. A formal email exercise in AP Spanish Language is:

• A linguistic skill → interpersonal writing at CEFR level B2 (the vertical dimension)
• A life competency → the ability to communicate effectively in writing, which is CCL in LOMLOE and EU_CCL in the European framework (the horizontal dimension)

These two dimensions are orthogonal: they never replace each other and are stored separately in every spec’s metadata.

VERTICAL DIMENSION                         HORIZONTAL DIMENSION
(subject area skills)                      (transversal competencies)
──────────────────────────────────────     ──────────────────────────────────────
What language skill does it practice?      What life/civic competency does it develop?

skill: interpersonal_writing               EU_CCL → Literacy competence
cefr:  B2                                  LOMLOE: CCL, CPSAA
bloom: APPLY                               IB Learner Profile: communicator
language: es                               OECD 2030: use_tools_interactively

In the OAS YAML spec, this produces two distinct metadata blocks:

metadata:
  competency_tags:               # ← VERTICAL dimension
    skills: [interpersonal_writing, formal_register]
    cefr:   [B2, C1]
    bloom:  [APPLY]
    language: es
    source_standard: AP_SLC

  transversal_competency_tags:   # ← HORIZONTAL dimension
    eu_key_competences: [EU_CCL]
    lomloe_cc:          [CCL, CPSAA]
    ib_learner_profile: [communicator, reflective]
    oecd_lc2030:        [use_tools_interactively]

Tip

Why two blocks? Vertical tags enable content discovery across standards (finding exercises with the same skill regardless of their country of origin). Horizontal tags enable compliance reporting to regulators (proving that a course covers LOMLOE’s Key Competences or IB’s Learner Profile attributes).

---

1. Subject-Area Competencies: The Vertical Dimension

Subject-area taxonomies define the universal skill vocabulary for a discipline. For Spanish language, this is anchored in CEFR (Common European Framework of Reference) and ACTFL Proficiency Guidelines — frameworks that apply globally regardless of whether the curriculum is AP, LOMLOE, IB, or SEP.

The Six Macro-Skills for Spanish

Skill ID	Name	CEFR Range	Standard Mappings
interpretive_reading	Reading Comprehension	A1–C2	AP Mode 1, LOMLOE CE.LCL.A, IB AO1
interpretive_listening	Listening Comprehension	A1–C2	AP Mode 1 (audio), LOMLOE CE.LCL.A.1
interpersonal_writing	Interactive Writing	A2–C1	AP FRQ Task 1, LOMLOE CE.LCL.B.1.1, IB Written Interaction
presentational_writing	Presentational Writing	A2–C2	AP FRQ Task 2, LOMLOE CE.LCL.B.1.4, IB Paper 2
interpersonal_speaking	Interactive Speaking	A2–C1	AP FRQ Task 3, LOMLOE CE.LCL.B.2.1, IB Oral
presentational_speaking	Presentational Speaking	A2–C2	AP FRQ Task 4, LOMLOE CE.LCL.B.2, IB Individual Oral

Beyond macro-skills, the taxonomy defines micro-competencies — fine-grained abilities that cut across multiple macro-skills:

Micro-Competency	Applies To	CEFR	Bloom
formal_register	Interpersonal & presentational writing/speaking	B1–C2	APPLY
argumentation	Presentational writing & speaking	B2–C2	ANALYZE, EVALUATE, CREATE
source_analysis	Writing & reading	B2–C2	ANALYZE, EVALUATE
cultural_comparison	Writing & speaking	B1–C2	ANALYZE, EVALUATE
literary_analysis	Reading & writing	B2–C2	ANALYZE, EVALUATE, CREATE

Note

Spanish as L1 (LOMLOE LCL, AP SLC, SEP) and Spanish as L2/FL (IB Spanish B) share the same skill vocabulary. The source_standard field in competency_tags indicates the pedagogical context. Spanish Literature (AP SLIT, IB Spanish A) uses a separate taxonomy file with extended literary analysis skills.

---

2. Transversal Competencies: The Horizontal Dimension

Transversal competencies apply across all subjects and countries. A student who writes a formal argumentative essay in Language class is simultaneously developing a competency that is equally relevant in History, Science, or Civics.

The Universal Base: EU Key Competences (2018)

The European Council’s 8 Key Competences serve as the universal anchor layer because:

• LOMLOE (Spain) adopted them almost 1:1 in its 8 Competencias Clave
• They are the reference framework for 40+ European countries
• They are compatible with IB Learner Profile, OECD Learning Compass, and P21

EU Key Competence          LOMLOE (ES)    IB Learner Profile    OECD LC 2030
─────────────────────────────────────────────────────────────────────────────
Literacy (CCL)         →   CCL        →  communicator        → use_tools_interactively
Multilingual (CP)      →   CP         →  communicator          —
Math+Science (STEM)    →   STEM       →  thinker             → critical thinking
Digital (CD)           →   CD         →  —                   → use_tools_interactively
Personal/Social (CPSAA)→   CPSAA      →  reflective, balanced → self_reflection
Citizenship (CC)       →   CC         →  principled           → act_with_responsibility
Entrepreneurship (CE)  →   CE         →  risk-taker           → create_new_value
Cultural Awareness(CCEC)→  CCEC       →  open-minded          —

Inference Rules

The platform’s AI agents automatically infer transversal competencies from the vertical skill classification — no manual tagging is needed:

If the exercise has…	→ Auto-assign
interpersonal_writing or presentational_writing	EU_CCL (CCL)
cultural_comparison or literary_analysis	EU_CCEC (CCEC)
argumentation with Bloom EVALUATE+	EU_CCL + EU_CPSAA
interpersonal_speaking	EU_CCL + EU_CP
Any formal production task	EU_CCL + EU_CPSAA

---

3. Cross-Standard Equivalencies: Sharing Content Across Countries

This is the feature that makes ColabEdu unique: the same pedagogical skill appears across AP, IB, LOMLOE, and EBAU under different names. The platform formally declares these equivalencies so that content created for one standard is automatically available to students from another.

The Five Universal Skill Bridges (Spanish)

Reading Comprehension

Skill ID: interpretive_reading · CEFR anchor: A1–C2

Standard	Task	Description
AP SLC	comp.ap.slc.interpretive	Interpretive Communication — Reading (Mode 1)
AP SLIT	comp.ap.slit.reading_multiple_choice	Section I Part A — Literary text comprehension
LOMLOE LCL	CE.LCL.A.2	Criterio de evaluación — Comprensión lectora
IB Spanish B	ib.sb.reading	Paper 1 — Comprehension (SL/HL)
IB Spanish A	ib.sa.paper1	Paper 1 — Guided literary analysis (unseen text)
EBAU	ebau.comprension_lectora	Pregunta de comprensión lectora

Any spec tagging these skills can be discovered by a LOMLOE pathway, an AP pathway, or an IB pathway — no manual mapping needed.

Argumentation

Skill ID: argumentation · CEFR anchor: B2–C2

Standard	Task	Description
AP SLC	comp.ap.slc.frq_task2	FRQ Task 2 — Argumentative essay with 3 sources
AP SLIT	comp.ap.slit.frq_task2	FRQ — Thematic essay comparing two literary texts
LOMLOE LCL	CE.LCL.B.1.4	Producción de textos argumentativos
IB Spanish B	ib.sb.writing	Paper 2 — Written task (argumentative/persuasive)
IB Spanish A	ib.sa.paper2	Paper 2 — Comparative essay with thesis
EBAU	ebau.redaccion_argumentativa	Redacción argumentativa

The AP FRQ Task 2, IB Paper 2, and EBAU essay are the same macro-competency expressed at different levels of formality. A student who masters one transfers it directly to the others.

Oral Interaction

Skill ID: interpersonal_speaking · CEFR anchor: A2–C1

Standard	Task	Description
AP SLC	comp.ap.slc.frq_task3	FRQ Task 3 — Simulated conversation (5 turns, 20 sec each)
LOMLOE LCL	CE.LCL.B.2.1	Interacción oral en situaciones formales e informales
IB Spanish B	ib.sb.oral	Individual oral — conversation with teacher (15 min)
IB Spanish A	ib.sa.ioa	Individual oral — literary and global issues (15 min, HL)

Critical gap for heritage speakers in AP SLC. Exercises from LOMLOE CE.LCL.B.2.1 and IB oral can fill this gap in an AP curriculum bundle.

Literary Analysis

Skill ID: literary_analysis · CEFR anchor: C1–C2

Standard	Task	Description
AP SLIT	comp.ap.slit.frq_task1	FRQ Task 1 — Short essay: textual analysis of one poem/prose
LOMLOE LCL	CE.LCL.C.1	Análisis de textos literarios
IB Spanish A	ib.sa.paper1	Paper 1 — Guided literary analysis (unseen text)
EBAU	ebau.comentario_literario	Comentario de texto literario guiado

The AP FRQ Task 1, LOMLOE CE.LCL.C.1, IB Paper 1, and EBAU commentary are equivalent in pedagogical competency. The difference is scope and source text, not the skill being assessed.

Cultural Comparison

Skill ID: cultural_comparison · CEFR anchor: B1–C2

Standard	Task	Description
AP SLC	comp.ap.slc.frq_task4	FRQ Task 4 — Cultural comparison (2 min oral presentation)
AP SLIT	comp.ap.slit.frq_task3	FRQ Task 3 — Cultural/historical analysis with visual art
LOMLOE LCL	CE.LCL.C.2	Comparación de textos de distintas épocas/culturas
IB Spanish B	ib.sb.oral	Individual oral — cultural comparison across two communities
IB Spanish A	ib.sa.ioa	Individual oral — global issues linked to literary text

---

4. The Taxonomy YAML Files

The entire competency system is declared in three types of YAML files inside the ce-specs repository. These files are the single source of truth consumed by all AI agents.

File Map

ce-specs/catalog/
│
├── global/
│   ├── taxonomy.competency_transversal.v1.yaml     ← HORIZONTAL layer
│   │   (EU Key Competences, national mappings,
│   │    inference rules, cross-standard equivalencies)
│   │
│   └── subject_areas/
│       ├── taxonomy.subject_area.spanish_language.v1.yaml   ← VERTICAL layer (L1/L2)
│       └── taxonomy.subject_area.spanish_literature.v1.yaml ← VERTICAL layer (Lit.)
│
├── es/
│   └── taxonomy.std.lomloe.yaml                    ← Spain-specific (Capa 2)
│
└── global/
    ├── taxonomy.std.ap.yaml                         ← AP-specific (Capa 2)
    └── taxonomy.ib_standards.v1.yaml                ← IB-specific (Capa 2)

What Each File Contains

taxonomy.competency_transversal.v1.yaml (horizontal, global)

• eu_key_competences — the 8 EU KC with mappings to LOMLOE CC, IB LP, OECD, P21
• national_mappings — how each country adapts the EU framework
• standard_task_rules — deterministic inference rules per task type (AP FRQ → EU KC)
• cross_standard_equivalencies — the 5 universal skill bridges with task_id for every standard

taxonomy.subject_area.spanish_language.v1.yaml (vertical, global)

• cefr_levels with ACTFL equivalencies
• skills (receptive, productive, interactive) with standard_mappings
• micro_competencies with Bloom levels
• classifier_rules — if/then rules for automatic skill inference from task metadata

taxonomy.std.ap.yaml / taxonomy.std.lomloe.yaml (standard-specific)

• The proprietary competency vocabulary of each program (AP FRQ Tasks, LOMLOE Criterios de Evaluación)
• Each task has a capa1_skills field that bridges to the universal skill vocabulary

Caution

Design rule: Subject-area taxonomy files (subject_areas/) never contain cross-standard equivalencies. Cross-standard equivalencies always live in taxonomy.competency_transversal.v1.yaml. Keeping these concerns separate is what makes the system extensible without touching every file when a new standard is added.

---

5. How AI Agents Use These Files

5.1 ACA / Curator — Ingestion Pipeline

The batch ingestion pipeline has a dedicated classify stage between parse and approval:

taxonomy → discover → fetch → parse → [classify] → approval → ingest

During classify, the CompetencyClassifier applies this sequence:

1. Reads source_standard from the spec generated by parse (e.g., AP_SPANISH_FRQ_TASK1)
2. Looks up capa1_skills in taxonomy.std.ap.yaml → [interpersonal_writing, formal_register], CEFR [B2, C1]
3. Confirms with classifier_rules in taxonomy.subject_area.spanish_language.v1.yaml
4. Infers transversal tags from standard_task_rules in taxonomy.competency_transversal.v1.yaml → EU_CCL, CPSAA
5. Writes competency_tags (vertical) and transversal_competency_tags (horizontal) into the spec’s metadata

This process is deterministic for known task types (no LLM involved). LLM inference is only used as a fallback for unknown task types.

5.2 Exercise Creation Wizard

When a teacher creates an exercise and selects an evaluation criterion (e.g., AP FRQ Task 2), the wizard automatically:

1. Loads the corresponding standard taxonomy ([C] files) to display criterion details
2. Calls the CompetencyClassifier to tag the generated exercise with vertical + horizontal competencies
3. Uses cross_standard_equivalencies to suggest rubrics from equivalent tasks in other standards

The result: a teacher using AP Spanish creates an exercise that is automatically available to a LOMLOE or IB curriculum bundle — without any extra effort.

5.3 Grading Agent

The grading pipeline uses competency tags in two ways:

Rubric selection: The GradingOrchestrator selects rubrics not only by standard but by competency_tags. If an AP FRQ Task 3 exercise exists, it can be evaluated with a LOMLOE CE.LCL.B.2.1 rubric (same interpersonal_speaking skill, same CEFR B2) if that rubric is richer or more appropriate.

Coverage reporting: After grading, the system aggregates transversal_competency_tags to generate competency coverage dashboards — showing which EU Key Competences and LOMLOE CC a student has developed across all their exercises, regardless of subject.

Student completes 20 exercises → System reports:
  EU_CCL covered:  15 exercises (75%)   ← Communication competence
  EU_CCEC covered:  5 exercises (25%)   ← Cultural awareness
  CPSAA covered:    8 exercises (40%)   ← Personal / learning to learn
  → Recommendation: add cultural comparison exercises (EU_CCEC gap)

---

6. Practical Impact: Content Reuse Across Countries

The cross-standard equivalency system unlocks a concrete economic and pedagogical benefit: exercises created for one program can be reused by another without manual curation.

End-to-End Flow: From AP Spec to LOMLOE Bundle

[Author creates AP FRQ Task 2 exercise]
         ↓
[CompetencyClassifier — deterministic, no LLM]
  reads:    taxonomy.std.ap.yaml → capa1_skills
  confirms: taxonomy.subject_area.spanish_language.v1.yaml
  infers:   taxonomy.competency_transversal.v1.yaml
         ↓
[Spec stored with full metadata]
  competency_tags:             ← VERTICAL
    skills: [presentational_writing, argumentation, source_analysis]
    cefr:   [C1]
    bloom:  [ANALYZE, EVALUATE]
  transversal_competency_tags: ← HORIZONTAL
    eu_key_competences: [EU_CCL, EU_CPSAA]
    lomloe_cc:          [CCL, CPSAA]
         ↓
[BundleAssembler LOMLOE looks for gap CE.LCL.B.1.4]
  competency_query: {skills: [argumentation], cefr: [C1]}
  → returns the AP exercise as a valid candidate     ✓
         ↓
[GradingOrchestrator grades with LOMLOE rubric]
  → uses the original CE.LCL.B.1.4 rubric            ✓
  → tags do not alter the grading                    ✓

How the BundleAssembler Queries Cross-Standard

A LOMLOE curriculum bundle requesting 20 reading comprehension exercises uses a competency_query instead of a program-specific token:

# LOMLOE pathway requesting reading comprehension exercises
content_targets:
  - spec_type: BLOCK_CONTENT
    count: 20
    competency_query:
      skills: [interpretive_reading]
      cefr:   [B2, C1]
      language: es

This query runs against a GIN-indexed JSONB column in PostgreSQL and returns exercises tagged with interpretive_reading + B2/C1 — regardless of whether they were originally created for AP, LOMLOE, IB, or EBAU. The system has no country bias.

Concrete Reuse Scenarios

Scenario	Origin standard	Destination standard	Bridge competency
Heritage speakers in AP SLC	LOMLOE / IB Oral	AP FRQ Task 3	interpersonal_speaking B2
EBAU bundle with AP material	AP FRQ Task 2	EBAU essay	argumentation C1
IB course extended with LOMLOE	LOMLOE CE.LCL.C.1	IB Spanish A Paper 1	literary_analysis C1–C2
EU dashboard for Spanish schools	Any exercise	EU Key Competences	All via transversal_competency_tags

Adding a New Standard

Extending the system to a new country (e.g., Brazil with BNCC or Mexico with SEP) requires changes to only three files:

1. Create br/taxonomy.std.bncc.yaml with BNCC competencies and capa1_skills bridges
2. Add standard_mappings for BNCC in taxonomy.subject_area.portuguese_language.v1.yaml (a few lines)
3. Add standard_task_rules for BNCC task types in taxonomy.competency_transversal.v1.yaml

No code changes required. The ACA, BundleAssembler, and GradingOrchestrator support the new standard automatically.

---

7. Related Documents

Document	Purpose
Language & Literature Rubrics	LOMLOE C0 rubric blocks for ESO Spanish Language and Literature
Educational Standards Crosswalk	Mapping table: how educational concepts translate across AP, IB, LOMLOE, SEP
OAS YAML Reference	Full schema reference for spec YAML files including competency_tags
Specs Repository	Overview of the ce-specs GitOps catalog