ThomasMoreAI/legal-skills-open / br/litigation/skills/juriscan-brunobracaioli
juriscan-brunobracaioli
VerifiedAnálise forense de processos judiciais brasileiros. Extrai, classifica e cruza peças processuais, detecta contradições, calcula prazos CPC e gera relatórios de risco com exportação para Obsidian. Use quando mencionar: processo judicial, petição, sentença, acórdão, contradições, prazos, timeline, análise forense, Obsidian.
tests/test_integrity_check.py
View on GitHub"""Tests for scripts/integrity_check.py"""
import sys
import os
import pytest
sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', 'scripts'))
from integrity_check import (
calculate_ocr_confidence,
detect_metadata_anomalies,
detect_page_gaps,
generate_integrity_report,
)
class TestOcrConfidence:
def test_clean_legal_text(self):
text = (
"SENTENÇA. Vistos. O autor ajuizou ação de indenização por danos "
"morais e materiais em face do réu, alegando que sofreu prejuízo "
"decorrente de vazamento de dados. A contestação foi apresentada "
"tempestivamente. O juiz de direito da comarca proferiu decisão."
)
score = calculate_ocr_confidence(text)
assert score >= 0.6
def test_garbled_text(self):
text = "xkzq bvnm rthl wcjf plmk sdfg hjkl zxcv bnmq wert yuip asdf ghjk"
score = calculate_ocr_confidence(text)
assert score < 0.4
def test_empty_text(self):
assert calculate_ocr_confidence('') == 0.0
def test_short_text(self):
assert calculate_ocr_confidence('abc') == 0.0
def test_mixed_quality(self):
text = (
"O processo foi distribuído em 2025. xkzqm bvnmt rthlp. "
"A sentença condenou o réu ao pagamento. wcjfx plmks sdfgy."
)
score = calculate_ocr_confidence(text)
# Should be between garbled and clean
assert 0.2 < score < 0.8
class TestMetadataAnomalies:
def test_no_anomalies(self):
chunks = [
{'label': 'PETIÇÃO INICIAL', 'primary_date': '05/02/2025'},
{'label': 'CONTESTAÇÃO', 'primary_date': '10/03/2025'},
{'label': 'SENTENÇA', 'primary_date': '01/07/2025'},
]
anomalies = detect_metadata_anomalies(chunks)
assert len(anomalies) == 0
def test_missing_contestacao(self):
chunks = [
{'label': 'PETIÇÃO INICIAL', 'primary_date': '05/02/2025'},
{'label': 'SENTENÇA', 'primary_date': '01/07/2025'},
]
anomalies = detect_metadata_anomalies(chunks)
faltante = [a for a in anomalies if a['tipo'] == 'PECA_FALTANTE']
assert len(faltante) == 1
def test_duplicate_sentenca(self):
chunks = [
{'label': 'PETIÇÃO INICIAL'},
{'label': 'SENTENÇA'},
{'label': 'SENTENÇA'},
]
anomalies = detect_metadata_anomalies(chunks)
dupes = [a for a in anomalies if a['tipo'] == 'PECA_DUPLICADA']
assert len(dupes) == 1
def test_duplicate_despacho_ok(self):
# Multiple DESPACHO is normal
chunks = [
{'label': 'DESPACHO'},
{'label': 'DESPACHO'},
{'label': 'DESPACHO'},
]
anomalies = detect_metadata_anomalies(chunks)
dupes = [a for a in anomalies if a['tipo'] == 'PECA_DUPLICADA']
assert len(dupes) == 0
def test_multiple_process_numbers(self):
chunks = [
{'label': 'PETIÇÃO INICIAL', 'processo_number': '1234567-89.2025.8.26.0100'},
{'label': 'CONTESTAÇÃO', 'processo_number': '9999999-00.2024.8.19.0001'},
]
anomalies = detect_metadata_anomalies(chunks)
divergent = [a for a in anomalies if a['tipo'] == 'NUMERO_PROCESSO_DIVERGENTE']
assert len(divergent) == 1
class TestPageGaps:
def test_no_gaps(self):
text = "fls. 1 ... fls. 2 ... fls. 3 ... fls. 4 ... fls. 5"
gaps = detect_page_gaps(text, {'pages': 10})
assert len(gaps) == 0
def test_large_gap(self):
text = "fls. 1 ... fls. 2 ... fls. 3 ... fls. 50 ... fls. 51"
gaps = detect_page_gaps(text, {'pages': 60})
lacunas = [g for g in gaps if g['tipo'] == 'LACUNA_PAGINAS']
assert len(lacunas) == 1
def test_incompatible_numbering(self):
text = "fls. 1 ... fls. 5 ... fls. 10 ... fls. 500"
gaps = detect_page_gaps(text, {'pages': 50})
incompat = [g for g in gaps if g['tipo'] == 'NUMERACAO_INCOMPATIVEL']
assert len(incompat) == 1
def test_too_few_markers(self):
text = "fls. 1 ... fls. 100"
gaps = detect_page_gaps(text, {'pages': 100})
assert len(gaps) == 0 # Not enough data points
class TestIntegrityReport:
def test_generates_report(self):
chunks = [
{
'index': 0,
'label': 'PETIÇÃO INICIAL',
'text': 'O autor ajuizou ação de indenização por danos morais perante o juiz da vara cível.',
'primary_date': '05/02/2025',
},
]
report = generate_integrity_report(chunks)
assert 'overall_confidence' in report
assert 'ocr_scores' in report
assert 'anomalies' in report
assert 'page_gaps' in report
assert len(report['ocr_scores']) == 1
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |