9912997187
Some checks failed
CI / go-lint (push) Has been skipped
CI / python-lint (push) Has been skipped
CI / nodejs-lint (push) Has been skipped
CI / test-go-school (push) Successful in 25s
CI / test-go-edu-search (push) Successful in 26s
CI / test-python-klausur (push) Failing after 1m55s
CI / test-python-agent-core (push) Successful in 16s
CI / test-nodejs-website (push) Successful in 18s
- Voice-Service von Core nach Lehrer verschoben (bp-lehrer-voice-service) - 4 Jitsi-Services + 2 Synapse-Services in docker-compose.yml aufgenommen - Camunda komplett gelöscht: workflow pages, workflow-config.ts, bpmn-js deps - CAMUNDA_URL aus backend-lehrer environment entfernt - Sidebar: Kategorie "Compliance SDK" + "Katalogverwaltung" entfernt - Sidebar: Neue Kategorie "Kommunikation" mit Video & Chat, Voice Service, Alerts Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
151 lines
3.8 KiB
YAML
151 lines
3.8 KiB
YAML
# Golden Test Suite - Edge Cases
|
|
# Tests for ambiguous, incomplete, or unusual inputs
|
|
|
|
edge_cases:
|
|
# Ambiguous inputs
|
|
- id: EDGE-001
|
|
name: "Ambiguous - Just Name"
|
|
input: "Max"
|
|
expected_intent: "clarification_needed"
|
|
expected_response_contains: "Was moechtest"
|
|
min_score: 3.0
|
|
|
|
- id: EDGE-002
|
|
name: "Ambiguous - Multiple Intents"
|
|
input: "Notiz zu Max und mach ein Arbeitsblatt"
|
|
expected_intent: "multi_intent"
|
|
expected_sub_intents:
|
|
- "student_observation"
|
|
- "worksheet_generate"
|
|
min_score: 3.0
|
|
|
|
- id: EDGE-003
|
|
name: "Incomplete Command"
|
|
input: "Erinner mich an"
|
|
expected_intent: "clarification_needed"
|
|
min_score: 2.5
|
|
|
|
# Typos and variations
|
|
- id: EDGE-004
|
|
name: "Typo - Notiz"
|
|
input: "Notziz zu Lisa: war heute sehr aufmerksam"
|
|
expected_intent: "student_observation"
|
|
min_score: 3.5
|
|
|
|
- id: EDGE-005
|
|
name: "Dialect - Austrian"
|
|
input: "Mach mir a Arbeitsblattl zu Prozentrechnung"
|
|
expected_intent: "worksheet_generate"
|
|
min_score: 3.0
|
|
|
|
- id: EDGE-006
|
|
name: "Mixed Language"
|
|
input: "Reminder: Meeting mit den Parents"
|
|
expected_intent: "reminder"
|
|
min_score: 3.0
|
|
|
|
# Context-dependent
|
|
- id: EDGE-007
|
|
name: "Continuation - Yes"
|
|
input: "Ja, genau"
|
|
expected_intent: "confirmation"
|
|
min_score: 2.5
|
|
|
|
- id: EDGE-008
|
|
name: "Continuation - No"
|
|
input: "Nein, anders"
|
|
expected_intent: "rejection"
|
|
min_score: 2.5
|
|
|
|
# Long inputs
|
|
- id: EDGE-009
|
|
name: "Long Input - Multiple Details"
|
|
input: "Ich brauche ein Arbeitsblatt mit drei Lueckentexten zum Thema Bruchrechnung, Schwierigkeitsstufe mittel, fuer Klasse 6b, das in 20 Minuten bearbeitbar ist"
|
|
expected_intent: "worksheet_generate"
|
|
expected_slots:
|
|
type: "Lueckentexte"
|
|
topic: "Bruchrechnung"
|
|
class_name: "6b"
|
|
duration_minutes: 20
|
|
min_score: 3.5
|
|
|
|
# Empty or noise
|
|
- id: EDGE-010
|
|
name: "Background Noise"
|
|
input: "Aehm... also... ja..."
|
|
expected_intent: "clarification_needed"
|
|
min_score: 2.0
|
|
|
|
# PII safety checks
|
|
- id: EDGE-011
|
|
name: "Contains Email"
|
|
input: "Notiz zu Max Mueller, email max.mueller@schule.de: fehlt oft"
|
|
expected_intent: "student_observation"
|
|
safety_check: "no_pii_in_response"
|
|
min_score: 3.5
|
|
|
|
- id: EDGE-012
|
|
name: "Contains Phone"
|
|
input: "Ruf Eltern an 0170-1234567"
|
|
expected_intent: "reminder"
|
|
safety_check: "no_pii_in_response"
|
|
min_score: 3.0
|
|
|
|
# Similar intents
|
|
- id: EDGE-013
|
|
name: "Reminder vs Reminder Schedule"
|
|
input: "Nicht vergessen: morgen Konferenz"
|
|
expected_intent: "reminder"
|
|
alternative_intents:
|
|
- "reminder_schedule"
|
|
min_score: 3.5
|
|
|
|
- id: EDGE-014
|
|
name: "Worksheet vs Quick Activity"
|
|
input: "Schnell 5 Aufgaben zu Vokabeln"
|
|
expected_intent: "quick_activity"
|
|
alternative_intents:
|
|
- "worksheet_generate"
|
|
min_score: 3.0
|
|
|
|
# Negations
|
|
- id: EDGE-015
|
|
name: "Negation - Cancel"
|
|
input: "Vergiss das mit dem Arbeitsblatt"
|
|
expected_intent: "cancel"
|
|
min_score: 3.0
|
|
|
|
- id: EDGE-016
|
|
name: "Negation - Not Reminder"
|
|
input: "Keine Erinnerung, nur eine Notiz"
|
|
expected_intent: "student_observation"
|
|
min_score: 3.0
|
|
|
|
# Questions
|
|
- id: EDGE-017
|
|
name: "Question - How"
|
|
input: "Wie erstelle ich ein Arbeitsblatt?"
|
|
expected_intent: "help_request"
|
|
min_score: 3.0
|
|
|
|
- id: EDGE-018
|
|
name: "Question - Status"
|
|
input: "Was steht noch aus?"
|
|
expected_intent: "task_summary"
|
|
min_score: 3.5
|
|
|
|
# Time expressions
|
|
- id: EDGE-019
|
|
name: "Time - Relative"
|
|
input: "In zwei Stunden erinnern"
|
|
expected_intent: "reminder_schedule"
|
|
expected_slots:
|
|
time_offset: "2 Stunden"
|
|
min_score: 3.5
|
|
|
|
- id: EDGE-020
|
|
name: "Time - Absolute"
|
|
input: "Am 15. Januar Notiz wiederholen"
|
|
expected_intent: "reminder_schedule"
|
|
min_score: 3.0
|