diff --git a/README.md b/README.md
index 166dcf6..087123d 100644
--- a/README.md
+++ b/README.md
@@ -1527,6 +1527,13 @@ aby dlhý prvý profil nevytlačil ostatné. Skrátenie rešpektuje celé odseky
 zoznamu vrátane ich výhrad; pri stručnom opise má zachovanie podmienky prednosť
 pred orientačným limitom dĺžky.
 
+Krátke porovnanie zobrazuje rovnaké oblasti vedľa seba v tabuľke; dlhé výňatky,
+kód a odkazy zostávajú v samostatných sekciách bez úpravy ich obsahu. Prázdne
+pomocné sekcie sa nevypisujú a chýbajúce oblasti sa konkrétne pomenujú pri danej
+osobe. Spoločné vlastnosti sa uvádzajú najprv súhrnne a potom s doloženými
+osobami a zdrojmi. Stručnejšie formulácie nemenia rok práce na dátum jej
+začatia alebo dokončenia a ciele neprezentujú ako dosiahnuté výsledky.
+
 Rozpoznané, ale nepodporované operátory (napríklad interval rokov, vylučovanie
 alebo minimálny počet) vedú k `ambiguous` a konkrétnemu spresneniu. Rok práce
 sa nepoužije ako dôkaz dátumu jej začatia či obhajoby. Chybové odpovede odlišujú
diff --git a/test/test_answer_clarity.py b/test/test_answer_clarity.py
new file mode 100644
index 0000000..44c63ec
--- /dev/null
+++ b/test/test_answer_clarity.py
@@ -0,0 +1,158 @@
+"""Concise presentation must keep factual bindings, qualifications and gaps."""
+import copy
+import sqlite3
+
+import pytest
+
+from test_quality_block import corpus, prepare, complete
+from test_evidence_selection import update_eva, selection
+from tools.answer_evaluation.context_policy import answer_context_issues
+from tools.answer_evaluation.content_selection import facet_lines
+from tools.answer_evaluation.graph_context import graph_context
+from tools.answer_evaluation.presentation import comparison_table
+from tools.answer_evaluation.rag_evidence import rag_evidence_context
+from tools.answer_evaluation.render_answer import render_graph_answer
+from tools.answer_evaluation.verified_rag import generate_verified_rag, rag_answer_issues
+
+
+def test_missing_aspects_share_one_specific_notice_without_empty_sections(corpus):
+    update_eva(corpus,'Stav:\n- Implementovaný parser.')
+    q='Aké výsledky, problémy a plánované experimenty má diplomová práca Evy Novej?'
+    p=prepare(corpus,q); c=rag_evidence_context(q,p)
+    result=generate_verified_rag(q,p,complete(selection(c)))
+    assert result['accepted_for_delivery'] and result['answer_coverage']['status']=='partial'
+    text=result['answer']
+    assert 'Implementovaný parser.' in text
+    assert text.count('V poskytnutých podkladoch chýbajú údaje')==1
+    assert 'riešených problémoch, plánovaných úlohách a plánovaných experimentoch' in text
+    assert '**Riešené problémy:**' not in text and '**Plánované experimenty' not in text
+    assert 'Táto časť odpovede' not in text and '\n\n\n' not in text
+
+
+def test_short_comparison_aligns_people_and_different_results(corpus):
+    q='Porovnaj výsledky Evy Novej a Borisa Nového.'
+    p=prepare(corpus,q); c=rag_evidence_context(q,p); a=selection(c)
+    before=copy.deepcopy((a,c))
+    lines=facet_lines(a,c); text='\n'.join(lines)
+    assert text.startswith('| Oblasť | Eva Nová — diplomová práca (2024) | Boris Nový — bakalárska práca (2023) |')
+    row=next(line for line in lines if line.startswith('| Zaznamenané výsledky |'))
+    assert 'BERT' in row.split('|')[2] and 'Python' in row.split('|')[3]
+    assert 'Konzultácia' not in text and 'Pripravte experiment' not in text
+    assert 'lepš' not in text and (a,c)==before
+    r=generate_verified_rag(q,p,complete(a))
+    assert r['accepted_for_delivery'] and len(r['source_urls'])==2
+
+
+@pytest.mark.parametrize('literal',[
+    'Implementovaný parser. Výsledok platí iba pre malý korpus. '+('Technický opis. '*70),
+    'Implementovaný parser pre `a | b` iba na testovacích dátach.',
+    'Implementovaný parser podľa https://example.test/method iba na testovacích dátach.',
+])
+def test_complex_comparisons_fall_back_without_losing_qualifiers_or_literals(corpus,literal):
+    update_eva(corpus,'Výsledky:\n- '+literal)
+    q='Porovnaj výsledky Evy Novej a Borisa Nového.'
+    p=prepare(corpus,q); c=rag_evidence_context(q,p); a=selection(c)
+    text='\n'.join(facet_lines(a,c))
+    assert not text.startswith('| Oblasť |') and literal.strip() in text
+    assert 'Boris Nový' in text and 'Python' in text
+    assert all('Ďalšie doložené' not in line for line in text.splitlines())
+
+
+def test_shared_excerpt_is_not_repeated_under_each_matching_facet(corpus):
+    update_eva(corpus,'Výsledky:\n- Vyhodnotená presnosť modelu v experimente.')
+    q='Aké experimenty a výsledky má diplomová práca Evy Novej?'
+    p=prepare(corpus,q); c=rag_evidence_context(q,p)
+    result=generate_verified_rag(q,p,complete(selection(c)))
+    assert result['accepted_for_delivery']
+    assert result['answer'].count('Vyhodnotená presnosť modelu v experimente.')==1
+    assert 'chýbajú údaje' not in result['answer'] and 'Doložené vyššie' not in result['answer']
+    assert 'Zaznamenaný výsledok:' in result['answer'] or '**Zaznamenané výsledky:**' in result['answer']
+
+
+def test_multiline_table_is_bounded_and_does_not_rewrite_source_html():
+    table=comparison_table(['Oblasť','Eva','Boris'],[('Výsledky',['Prvý výsledok.\nDruhý výsledok.','Parser.'])])
+    assert table and 'Prvý výsledok.<br>Druhý výsledok.' in table[-1]
+    assert comparison_table(['Oblasť','Eva'],[('Výsledky',['<b>Prvý výsledok.</b>'])]) is None
+
+
+def test_shared_result_precedes_people_and_citations_without_repeating_tags():
+    people=['Ján Pták','Ján Malinovský']; urls=['https://example.test/ptak','https://example.test/malinovsky']
+    claims=[{'person':person,'keyword':tag} for person in people for tag in ['NLP','RAG']]
+    c={'status':'ok','complete':True,'truncated':False,'facts':[
+        {**claim,'source_url':urls[people.index(claim['person'])]} for claim in claims],
+       'sources':[{'source_url':u} for u in urls], 'requested_people':people,
+       'shared_properties':[{'field':'keyword','value':tag,'people':people} for tag in ['NLP','RAG']]}
+    a={'status':'ok','claims':claims,'source_urls':urls}
+    r=render_graph_answer(a,c); text=r['answer']
+    assert text.startswith('Spoločné kľúčové slová: **NLP**, **RAG**.')
+    assert text.index('RAG')<text.index('Doložené pri:')<text.index('Zdroje:')
+    assert all(text.count('**'+tag+'**')==1 for tag in ['NLP','RAG'])
+    assert all(person in text for person in people)
+    assert not answer_context_issues(r,c)
+
+
+def test_negative_names_failed_year_but_does_not_license_positive_year_claims():
+    c={'status':'no_answer','facts':[],'sources':[],'requested_people':['Martin Šarišský'],
+       'requested_years':[1999],'complete':True}
+    a=render_graph_answer({'status':'no_answer','claims':[],'source_urls':[]},c)
+    assert 'Martin Šarišský' in a['answer'] and 'z roku **1999**' in a['answer']
+    assert not answer_context_issues(a,c)
+    for year in [1999,2099]:
+        bad={**a,'answer':a['answer']+f' Práca bola dokončená v roku {year}.'}
+        assert 'unsupported_prose_number' in answer_context_issues(bad,c)
+
+
+def test_start_year_is_not_rendered_as_a_missing_academic_work_year():
+    c=graph_context({'status':'no_answer','reason':'not_found','facts':[],'sources':[],'aggregation':None,
+        'complete':True,'constraints':{'years':[1999],'mentions_start_year':True},
+        'resolved_entities':[{'kind':'Person','name':'Martin Šarišský'}]})
+    a=render_graph_answer({'status':'no_answer','claims':[],'source_urls':[]},c)
+    assert 'záznam práce' not in a['answer'] and '1999' not in a['answer']
+
+
+def test_partial_overview_retains_supported_person_and_names_missing_description(corpus):
+    db,raw=corpus
+    with sqlite3.connect(db) as conn:
+        conn.execute("UPDATE chunks SET text='Názov: Ukážkový názov.' WHERE document_path LIKE '%boris_novy%'")
+    q='Porovnaj obsah diplomovej práce Evy Novej a bakalárskej práce Borisa Nového.'
+    p=prepare(corpus,q); c=rag_evidence_context(q,p)
+    assert not c['requirements'].get('facets')
+    passage=next(p for p in c['content_passages'] if p['person']=='Eva Nová')
+    a={'status':'ok','answer':'','claims':[], 'source_urls':[s['source_url'] for s in c['sources']],
+       'evidence_quotes':[{'source_id':passage['source_id'],'passage_id':passage['passage_id'],
+                          'quote':'Implementujte systém vyhľadávania.'}]}
+    assert not rag_answer_issues(a,c)
+    result=generate_verified_rag(q,p,complete(a))
+    assert result['accepted_for_delivery'] and result['answer_coverage']['status']=='partial'
+    assert 'Implementácia systému vyhľadávania' in result['answer']
+    assert 'Boris Nový: v poskytnutých podkladoch chýba opis požadovanej práce alebo projektu.' in result['answer']
+    assert 'Výpis je neúplný' not in result['answer']
+
+
+def test_available_description_cannot_be_silently_omitted(corpus):
+    q='Porovnaj obsah diplomovej práce Evy Novej a bakalárskej práce Borisa Nového.'
+    p=prepare(corpus,q); c=rag_evidence_context(q,p)
+    passage=next(p for p in c['content_passages'] if p['person']=='Eva Nová')
+    a={'status':'ok','answer':'','claims':[], 'source_urls':[s['source_url'] for s in c['sources']],
+       'evidence_quotes':[{'source_id':passage['source_id'],'passage_id':passage['passage_id'],
+                          'quote':'Implementujte systém vyhľadávania.'}]}
+    assert 'missing_bound_content_evidence' in rag_answer_issues(a,c)
+
+
+@pytest.mark.parametrize('aspect',['obsah','zameranie'])
+def test_overview_comparison_keeps_each_work_in_its_own_column(corpus,aspect):
+    q=f'Porovnaj {aspect} diplomovej práce Evy Novej a bakalárskej práce Borisa Nového.'
+    p=prepare(corpus,q); c=rag_evidence_context(q,p)
+    a={'status':'ok','answer':'','claims':[], 'source_urls':[s['source_url'] for s in c['sources']],
+       'evidence_quotes':[{'source_id':passage['source_id'],'passage_id':passage['passage_id'],'quote':text}
+            for person,text in [('Eva Nová','Implementujte systém vyhľadávania.'),
+                                ('Boris Nový','Implementujte parser.')]
+            for passage in c['content_passages'] if passage['person']==person and text in passage['text']]}
+    result=generate_verified_rag(q,p,complete(a))
+    assert result['accepted_for_delivery'] and result['answer_coverage']['status']=='supported'
+    assert result['answer'].startswith('| Oblasť | Diplomová práca Evy Novej (2024) | Bakalárska práca Borisa Nového (2023) |')
+    row=next(line for line in result['answer'].splitlines() if line.startswith('| Zameranie podľa podkladov |'))
+    assert 'vyhľadávania' in row.split('|')[2] and 'parser' in row.split('|')[3]
+    assert 'Pripravte experiment' not in row and len(result['source_urls'])==2
+    refused={**a,'status':'no_answer','evidence_quotes':[],'source_urls':[]}
+    assert 'rejected_supported_work_description' in rag_answer_issues(refused,c)
diff --git a/test/test_answer_presentation.py b/test/test_answer_presentation.py
index b3025dd..ff39ed8 100644
--- a/test/test_answer_presentation.py
+++ b/test/test_answer_presentation.py
@@ -17,8 +17,8 @@ def test_identity_is_a_sentence_but_year_is_not_a_start_or_completion_date():
              'title': 'Systém získavania informácií.', 'year': 2021}
     before = copy.deepcopy(claim)
     text = fact_sentences(claim, 'Aký je názov diplomovej práce Jána Holpa a z ktorého roku je?')
-    assert text == ('Diplomová práca Jána Holpa má názov **Systém získavania informácií**. '
-                    'Rok práce je **2021**.')
+    assert text == ('Diplomová práca Jána Holpa má názov **Systém získavania informácií** '
+                    '(rok práce: **2021**).')
     assert not any(word in text for word in ('začal', 'dokonč', 'vypracovaná'))
     assert claim == before
 
@@ -115,6 +115,6 @@ def test_negative_is_a_bounded_absence_not_a_claim_about_real_existence():
          'complete': True, 'question': 'Akú prácu mal Ján Holp v roku 1999?'}
     a = {'status': 'no_answer', 'claims': [], 'source_urls': []}
     text = render_graph_answer(a, c)['answer']
-    assert 'Ján Holp' in text and 'zodpovedajúce tejto otázke' in text
+    assert 'Ján Holp' in text and 'požadované údaje sa v dostupných podkladoch nenašli' in text
     assert 'neexistuje' not in text and '2021' not in text
     assert answer_context_issues(render_graph_answer(a, c), c) == []
diff --git a/test/test_comparison_answers.py b/test/test_comparison_answers.py
index f8c6763..8a9f56c 100644
--- a/test/test_comparison_answers.py
+++ b/test/test_comparison_answers.py
@@ -48,7 +48,7 @@ def test_goal_conversion_is_not_all_or_nothing():
         'Ján Malinovský', 'diploma_project', 'Aké sú ciele diplomového projektu Jána Malinovského?'))
     assert 'Diplomový projekt Jána Malinovského' in text
     assert 'Zistiť, čo je to RAG.' in text and 'Použiť databázu Neo4j.' in text
-    assert 'Ďalšie podklady:' in text and 'Práca neimplementuje vlastný model.' in text
+    assert 'Ďalšie podklady:' not in text and 'Práca neimplementuje vlastný model.' in text
     assert not any(word in text for word in ['Zistite', 'Naučte', 'Zostavte', 'Využite', 'Použite'])
 
 
@@ -150,7 +150,9 @@ def test_comparison_separates_goals_and_results(comparison_payload):
     result=generate_verified_rag(COMPARE,payload,lambda _: {'choices':[{'message':{'content':json.dumps(selected)}}]})
     assert result['accepted_for_delivery']
     text=result['answer']
-    assert text.count('**Zadanie a ciele:**')==text.count('**Zaznamenané výsledky:**')==2
+    assert text.startswith('| Oblasť |')
+    assert text.count('| Zadanie a ciele |')==text.count('| Zaznamenané výsledky |')==1
+    assert 'Diplomová práca Jána Holpa (2021)' in text and 'Bakalárska práca Ondreja Megelu (2020)' in text
     assert 'Natrénovať jazykový model metódou BERT.' in text
     assert 'Natrénovaný model wiki103' in text and 'Implementovaný PageRank' in text
     assert 'Cassandra' not in text and len(result['source_urls'])==2
@@ -172,7 +174,7 @@ def test_missing_results_allow_an_explicit_partial_comparison(comparison_payload
     selected=selected_comparison(c)
     result=generate_verified_rag(COMPARE,payload,lambda _: {'choices':[{'message':{'content':json.dumps(selected)}}]})
     assert result['accepted_for_delivery']
-    assert 'Táto časť odpovede nie je v poskytnutých podkladoch doložená.' in result['answer']
+    assert '| Zaznamenané výsledky | V podkladoch chýbajú údaje. |' in result['answer']
     assert 'Implementácia systému vyhľadávania' in result['answer']
 
 
diff --git a/test/test_quality_block.py b/test/test_quality_block.py
index 7d4d821..950ad57 100644
--- a/test/test_quality_block.py
+++ b/test/test_quality_block.py
@@ -162,7 +162,7 @@ def test_partial_answer_names_the_missing_aspect(corpus):
     q='Aké problémy riešili Eva Nová a Boris Nový a aké experimenty mali pripraviť?'
     p=prepare(corpus,q); r=generate_verified_rag(q,p,complete(model_selection(q,p)))
     assert r['accepted_for_delivery'] and r['answer_coverage']['status']=='partial'
-    assert 'Táto časť odpovede nie je v poskytnutých podkladoch doložená.' in r['answer']
+    assert 'V poskytnutých podkladoch chýbajú údaje o plánovaných experimentoch.' in r['answer']
     assert 'parserom' in r['answer']
 
 
diff --git a/test/test_rag_answer_coverage.py b/test/test_rag_answer_coverage.py
index 47c1bf3..fc269cd 100644
--- a/test/test_rag_answer_coverage.py
+++ b/test/test_rag_answer_coverage.py
@@ -113,7 +113,8 @@ def test_no_answer_is_bounded_and_only_reports_actual_context_cut(payload, budge
     c = rag_evidence_context('Aké je zadanie diplomovej práce Evy Novej v roku 1999?', payload)
     a = {'status': 'no_answer', 'claims': [], 'evidence_quotes': [], 'source_urls': []}
     text = render_rag_answer(a, c)
-    assert 'nepodarilo spoľahlivo nájsť' in text and 'Výpis je neúplný' not in text
+    assert 'Eva Nová' in text and 'v dostupných podkladoch sa nenašiel záznam diplomovej práce z roku **1999**' in text
+    assert 'Výpis je neúplný' not in text
     assert ('Časť podkladov bola pri príprave kontextu skrátená.' in text) == budget
     assert answer_coverage(a, c)['status'] == 'not_found'
 
diff --git a/tools/answer_evaluation/answer_coverage.py b/tools/answer_evaluation/answer_coverage.py
index c6b9eda..025beec 100644
--- a/tools/answer_evaluation/answer_coverage.py
+++ b/tools/answer_evaluation/answer_coverage.py
@@ -21,7 +21,7 @@ def response_scope(question: str, *, content: bool = False) -> str:
     if content:
         if re.search(r'\b(?:zadani\w*|ciel\w*)\b', query) and not re.search(r'\b(?:strucn\w*|zhrn\w*)\b', query):
             return 'assignment'
-        if re.search(r'\b(?:comu|o com|opis\w*|popis\w*|obsah\w*|zhrn\w*|strucn\w*)\b', query):
+        if re.search(r'\b(?:comu|o com|opis\w*|popis\w*|obsah\w*|zamer\w*|zhrn\w*|strucn\w*)\b', query):
             return 'overview'
     return 'fact'
 
@@ -55,6 +55,10 @@ def answer_coverage(answer: dict, context: dict) -> dict:
             from .content_selection import missing_facets
             if missing_facets(answer,context):
                 coverage='partial'; warnings.append('partial_requested_aspects')
+        elif requirements.get('content'):
+            from .rag_evidence import missing_content_people
+            if missing_content_people(answer,context):
+                coverage='partial'; warnings.append('partial_requested_aspects')
         if scope == 'list' and (limited or clipped):
             coverage = 'partial'; warnings.append('list_may_be_incomplete')
         elif scope == 'count' and (context.get('aggregation') or {}).get('complete') is not True:
diff --git a/tools/answer_evaluation/content_selection.py b/tools/answer_evaluation/content_selection.py
index 5ccb003..c753a66 100644
--- a/tools/answer_evaluation/content_selection.py
+++ b/tools/answer_evaluation/content_selection.py
@@ -218,34 +218,22 @@ def fit_candidates(sent: dict, context: dict, limit: int, encoding: str) -> None
 
 
 def facet_lines(answer: dict, context: dict) -> list[str]:
-    from .presentation import goal_text, work_subject, owner_in_question
-    from .rag_evidence import WORK_LABELS
+    from .presentation import goal_text, evidence_subject, missing_aspects, comparison_table
     req=context['requirements']; units=evidence_units(context,answer.get('evidence_quotes',[]))
-    lines=[]
+    facets=req.get('facets',[]);groups=[];labels={f:FACET_LABELS[f] for f in facets}
+    if 'experiments' in facets and planned_experiments(context.get('question','')):
+        labels['experiments']='Plánované experimenty a vyhodnotenie'
     for person in req['people']:
-        scopes={(p['work_type'],p.get('year')) for p in context.get('content_passages',[]) if p.get('person')==person}
-        kind=req['work_types'].get(person) or (next(iter(scopes))[0] if len(scopes)==1 else None)
-        subject=work_subject(person,kind,context.get('question',''))
-        if not owner_in_question(person,context.get('question','')):
-            subject=person+(' — '+WORK_LABELS[kind] if kind in WORK_LABELS else '')
-        if len(scopes)==1 and next(iter(scopes))[1] is not None:
-            subject+=' ('+str(next(iter(scopes))[1])+')'
-        labels={p.get('work_label') for p in context.get('content_passages',[])
-                if p.get('person')==person and p.get('work_label')}
-        if req.get('project_number') is not None and len(labels)==1:
-            subject=person+' — '+' '.join(next(iter(labels)).split())
-        lines.extend(['**'+subject+'**', ''])
-        seen=set()
-        for facet in req.get('facets', []):
-            label=FACET_LABELS[facet]
-            if facet=='experiments' and planned_experiments(context.get('question','')):
-                label='Plánované experimenty a vyhodnotenie'
+        group={'subject':evidence_subject(person,context),'items':{},'missing':[]};seen=set()
+        for facet in facets:
             rows=[u for u in units if u['person']==person and facet in u['facets']]
+            if facet=='progress' and 'results' in facets and any(u['phase']=='progress' for u in rows):
+                rows=[u for u in rows if u['phase']=='progress']
             rows.sort(key=lambda u:-unit_score(u['text'],u['kind'],facet))
-            lines.append('**'+label+':**')
-            shown=0
+            if not rows:group['missing'].append(facet)
+            rendered=[]
             for row in rows:
-                key=' '.join(re.sub(r'^\s*(?:\d+[.)]|[-+*])\s+','',row['text']).split())
+                key=(' '.join(re.sub(r'^\s*(?:\d+[.)]|[-+*])\s+','',row['text']).split()),row['phase'])
                 if key in seen:
                     continue
                 text=row['text']
@@ -253,14 +241,30 @@ def facet_lines(answer: dict, context: dict) -> list[str]:
                     text,_=goal_text(text)
                 text=re.sub(r'^\s*(?:\d+[.)]|[-+*])\s+','',text)
                 phase=' (plán)' if row['kind'] in {'assignment','task'} and facet not in {'assignment','plans'} else ''
-                state=({'result':'Zaznamenaný výsledok: ','progress':'Priebežný stav: '}.get(row['phase'],'')
-                       if facet=='results' else '')
-                lines.append('- '+state+text+phase); seen.add(key); shown+=1
-                if shown==2: break
-            if not shown:
-                if rows:
-                    lines.append('Je doložené rovnakým výňatkom uvedeným vyššie.')
-                else:
-                    lines.append('Táto časť odpovede nie je v poskytnutých podkladoch doložená.')
-            lines.append('')
+                state='Zaznamenaný výsledok: ' if row['phase']=='result' and 'results' in facets and facet!='results' else ''
+                rendered.append(state+text+phase);seen.add(key)
+                if len(rendered)==2:break
+            group['items'][facet]=rendered
+        groups.append(group)
+    if len(groups)>1 and re.search(r'\bporovnaj\w*\b',normalized(context.get('question',''))):
+        rows=[]
+        for facet in facets:
+            if not any(g['items'][facet] for g in groups):
+                if all(facet not in g['missing'] for g in groups):continue
+            values=['\n'.join(g['items'][facet]) or ('V podkladoch chýbajú údaje.' if facet in g['missing'] else 'Doložené vyššie.') for g in groups]
+            rows.append((labels[facet],values))
+        if any(any(g['items'][f] for f in facets) for g in groups):
+            table=comparison_table(['Oblasť',*(g['subject'] for g in groups)],rows)
+            if table is not None:return table
+    lines=[]
+    for group in groups:
+        title=group['subject']
+        if len(facets)==1 and group['items'][facets[0]]:title+=' — '+labels[facets[0]].lower()
+        lines.extend(['**'+title+'**',''])
+        for facet in facets:
+            items=group['items'][facet]
+            if not items:continue
+            if len(facets)>1:lines.append('**'+labels[facet]+':**')
+            lines.extend('- '+text for text in items);lines.append('')
+        if group['missing']:lines.extend([missing_aspects(group['missing'],context.get('question','')),''])
     return lines
diff --git a/tools/answer_evaluation/context_policy.py b/tools/answer_evaluation/context_policy.py
index 54b98f9..b9f3afd 100644
--- a/tools/answer_evaluation/context_policy.py
+++ b/tools/answer_evaluation/context_policy.py
@@ -108,9 +108,15 @@ def coverage_issues(answer: dict, context: dict) -> list[str]:
     return list(dict.fromkeys(issues))
 
 
-def prose_issues(answer: dict) -> list[str]:
+def prose_issues(answer: dict, context: dict | None = None) -> list[str]:
     """Conservative numeric consistency check; final prose is rendered from claims."""
     text = str(answer.get('answer', ''))
+    if context is not None and answer.get('status')=='no_answer':
+        from .presentation import no_answer_text
+        expected=no_answer_text(context)
+        # These years describe the failed query, never a positive factual claim.
+        # Strip only the exact trusted template; appended prose stays checked.
+        if text.startswith(expected):text=text[len(expected):]
     for url in sorted(text_urls(text), key=len, reverse=True):
         text = text.replace(url, '')
     # A bounded ordered-list marker is presentation, not a factual count/year.
@@ -236,7 +242,7 @@ def answer_context_issues(answer: dict, context: dict, *, check_text: bool = Tru
             issues.append('split_or_missing_person_join')
     if context.get('status') in {'no_answer','ambiguous','error'} and answer.get('status')=='ok':
         issues.append('unsupported_positive_status')
-    issues.extend(prose_issues(answer))
+    issues.extend(prose_issues(answer,context))
     if check_text or str(answer.get('answer', '')).strip():
         issues.extend(coverage_issues(answer,context))
     return list(dict.fromkeys(issues))
diff --git a/tools/answer_evaluation/graph_context.py b/tools/answer_evaluation/graph_context.py
index 3b64fe9..b4b3e3a 100644
--- a/tools/answer_evaluation/graph_context.py
+++ b/tools/answer_evaluation/graph_context.py
@@ -61,6 +61,8 @@ def graph_context(result: dict) -> dict:
              'index_version':result.get('index_version')}
     context['answer_candidates']=joined_candidates(facts,intent)
     context['question'] = result.get('query', '')
+    context['requested_years']=intent.get('years',[]) if not intent.get('mentions_start_year') else []
+    context['requested_work_types']=[*intent.get('thesis_types',[]),*intent.get('work_types',[])]
     context['clarification'] = result.get('clarification')
     context['requested_people'] = list(dict.fromkeys(row['name']
         for row in result.get('resolved_entities', []) if row.get('kind') == 'Person' and row.get('name')))
diff --git a/tools/answer_evaluation/presentation.py b/tools/answer_evaluation/presentation.py
index e7d8161..9501e14 100644
--- a/tools/answer_evaluation/presentation.py
+++ b/tools/answer_evaluation/presentation.py
@@ -53,7 +53,8 @@ def fact_sentences(claim: dict, question: str = '') -> str:
         if owner:
             subject = work_subject(person, kind, question)
             if title is not None:
-                parts.append(f'{subject} má názov **{displayed_title}**.')
+                parts.append(f'{subject} má názov **{displayed_title}**'+
+                             (f' (rok práce: **{year}**).' if year is not None else '.'))
             elif year is not None:
                 parts.append(f'{subject} je v podkladoch evidovaná s rokom **{year}**.'
                              if not (kind or '').endswith('_project') else
@@ -69,14 +70,17 @@ def fact_sentences(claim: dict, question: str = '') -> str:
             if year is not None:
                 sentence += f' (rok práce: **{year}**)'
             parts.append(sentence + '.')
-        if owner and title is not None and year is not None:
-            parts.append(f'Rok práce je **{year}**.')
     if claim.get('start_year') is not None:
         parts.append(f'{person} má uvedený začiatok štúdia v roku **{claim["start_year"]}**.')
     for key, label in [('keyword', 'kľúčové slovo'), ('category', 'kategória'),
                        ('document_author', 'autor dokumentu')]:
         if claim.get(key) is not None:
-            parts.append(f'Pri osobe {person} je uvedené: {label} **{claim[key]}**.')
+            if key=='keyword':
+                if parts and (kind or title is not None or year is not None):
+                    parts[-1]=parts[-1].rstrip('.')+f' a tag **{claim[key]}**.'
+                else:parts.append(f'{person} má tag **{claim[key]}**.')
+            else:
+                parts.append(f'{person}: '+('autor dokumentu je' if key=='document_author' else 'uvedená kategória je')+f' **{claim[key]}**.')
     return ' '.join(parts) or person + '.'
 
 
@@ -158,7 +162,7 @@ def content_lines(units: list[str], person: str | None, kind: str | None,
         lines = [subject + ' má podľa dostupných podkladov tieto ciele:', '',
                  *[f'{index}. {text}' for index, text in enumerate(goals, 1)]]
         if other:
-            lines.extend(['', 'Ďalšie podklady:', *other])
+            lines.extend(['', *other])
         return lines
     return [subject + ' je v dostupných podkladoch opísaná takto:' if not (kind or '').endswith('_project')
             else subject + ' je v dostupných podkladoch opísaný takto:', '', *units]
@@ -167,6 +171,62 @@ def content_lines(units: list[str], person: str | None, kind: str | None,
 def no_answer_text(context: dict) -> str:
     people = context.get('requested_people') or context.get('requirements', {}).get('people', [])
     if len(people) == 1:
-        return (f'Pre osobu {people[0]} sa v dostupných podkladoch nepodarilo spoľahlivo nájsť '
-                'údaje zodpovedajúce tejto otázke.')
-    return 'K tejto otázke sa v dostupných podkladoch nepodarilo spoľahlivo nájsť zodpovedajúce údaje.'
+        req=context.get('requirements',{})
+        years=context.get('requested_years') or req.get('years') or []
+        kinds=context.get('requested_work_types') or [req.get('work_types',{}).get(people[0])]
+        genitive={'bachelor_thesis':'bakalárskej práce','diploma_thesis':'diplomovej práce',
+                  'doctoral_thesis':'dizertačnej práce','bachelor_project':'bakalárskeho projektu',
+                  'diploma_project':'diplomového projektu','team_project':'tímového projektu',
+                  'research_project':'výskumného projektu'}
+        kind=genitive.get(kinds[0],'práce alebo projektu') if len(kinds)==1 else 'práce alebo projektu'
+        if years:
+            date=(' z roku **'+str(years[0])+'**' if len(years)==1 else
+                  ' z požadovaných rokov **'+', '.join(map(str,years))+'**')
+            return f'{people[0]}: v dostupných podkladoch sa nenašiel záznam {kind}{date}.'
+        return f'{people[0]}: požadované údaje sa v dostupných podkladoch nenašli.'
+    return 'Požadované údaje sa v dostupných podkladoch nenašli.'
+
+
+def missing_aspects(facets: list[str], question: str = '') -> str:
+    """Name only the missing requested fields, without empty helper sections."""
+    from .question_intent import planned_experiments
+    labels={'assignment':'zadaní a cieľoch','plans':'plánovaných úlohách','progress':'priebežnom stave',
+            'results':'výsledkoch','problems':'riešených problémoch','methods':'metódach a technológiách',
+            'experiments':('plánovaných experimentoch' if planned_experiments(question) else 'experimentoch a ich vyhodnotení')}
+    values=list(dict.fromkeys(labels.get(f,f) for f in facets))
+    if not values:return ''
+    joined=values[0] if len(values)==1 else ', '.join(values[:-1])+' a '+values[-1]
+    return 'V poskytnutých podkladoch chýbajú údaje o '+joined+'.'
+
+
+def clean_answer(lines: list[str]) -> str:
+    text='\n'.join(lines).strip()
+    return re.sub(r'\n{3,}','\n\n',text)
+
+
+def evidence_subject(person: str, context: dict) -> str:
+    """Use only work identity actually present in the supplied passages."""
+    req=context.get('requirements',{})
+    passages=[p for p in context.get('content_passages',[]) if p.get('person')==person]
+    scopes={(p.get('work_type'),p.get('year')) for p in passages}
+    if not scopes:return person
+    kind=req.get('work_types',{}).get(person) or (next(iter(scopes))[0] if len(scopes)==1 else None)
+    subject=work_subject(person,kind,context.get('question',''))
+    if not owner_in_question(person,context.get('question','')):
+        subject=person+(' — '+WORK_LABELS[kind] if kind in WORK_LABELS else '')
+    if len(scopes)==1 and next(iter(scopes))[1] is not None:
+        subject+=' ('+str(next(iter(scopes))[1])+')'
+    labels={p.get('work_label') for p in passages if p.get('work_label')}
+    if req.get('project_number') is not None and len(labels)==1:
+        subject=person+' — '+' '.join(next(iter(labels)).split())
+    return subject
+
+
+def comparison_table(headers: list[str], rows: list[tuple[str,list[str]]]) -> list[str] | None:
+    """Lay out short literal evidence; keep complex text outside Markdown tables."""
+    cells=[*headers,*(cell for label,values in rows for cell in [label,*values])]
+    if len(headers)>4 or any(len(cell)>900 or '|' in cell or '`' in cell or
+            re.search(r'<[^>]+>',cell) or text_urls(cell) for cell in cells):
+        return None
+    def row(values):return '| '+' | '.join(v.replace('\n','<br>') for v in values)+' |'
+    return [row(headers),row(['---']*len(headers)),*(row([label,*values]) for label,values in rows)]
diff --git a/tools/answer_evaluation/rag_evidence.py b/tools/answer_evaluation/rag_evidence.py
index 433bfea..411233e 100644
--- a/tools/answer_evaluation/rag_evidence.py
+++ b/tools/answer_evaluation/rag_evidence.py
@@ -280,6 +280,14 @@ def quote_matches_passage(quote: dict, passage: dict) -> bool:
             ' '.join(quote['quote'].split()) in ' '.join(passage['text'].split()))
 
 
+def missing_content_people(answer: dict, context: dict) -> list[str]:
+    """Requested people without a selected, work-bound descriptive excerpt."""
+    return [person for person in context.get('requirements',{}).get('people',[])
+            if not any(p.get('person')==person and quote_matches_passage(q,p)
+                       for p in context.get('content_passages',[])
+                       for q in answer.get('evidence_quotes',[]))]
+
+
 def quote_sections(quote: dict, context: dict) -> list[dict]:
     """Classify only literal excerpts bound to the delivered academic passage."""
     full = [section for passage in context.get('content_passages', [])
diff --git a/tools/answer_evaluation/render_answer.py b/tools/answer_evaluation/render_answer.py
index b724015..e2dc034 100644
--- a/tools/answer_evaluation/render_answer.py
+++ b/tools/answer_evaluation/render_answer.py
@@ -3,7 +3,7 @@ from copy import deepcopy
 import re
 
 from tools.graphrag.graphrag import WORK_TYPE_LABELS
-from .presentation import fact_sentences, no_answer_text
+from .presentation import fact_sentences, no_answer_text, clean_answer
 
 
 def render_graph_answer(answer: dict, context: dict) -> dict:
@@ -11,6 +11,7 @@ def render_graph_answer(answer: dict, context: dict) -> dict:
     lines = []
     if out['status'] == 'ok':
         shared_fields=set()
+        shared_people=[]
         labels={'keyword':'kľúčové slová','category':'kategórie','document_author':'autori dokumentov'}
         for field,label in labels.items():
             values=sorted({row['value'] for row in context.get('shared_properties',[]) if row['field']==field
@@ -20,6 +21,11 @@ def render_graph_answer(answer: dict, context: dict) -> dict:
                 shared_fields.add(field)
                 prefix='Spoločné ' if context.get('complete') is True and not context.get('truncated') else 'Doložené spoločné '
                 lines.append(prefix+label+': '+', '.join('**'+v+'**' for v in values)+'.')
+                for row in context.get('shared_properties',[]):
+                    if row['field']==field and row['value'] in values:
+                        shared_people.extend(row['people'])
+        if shared_people:
+            lines.extend(['Doložené pri: '+', '.join('**'+p+'**' for p in dict.fromkeys(shared_people))+'.',''])
         comparison = context.get('academic_year_comparison')
         if comparison:
             people = comparison.get('people', [])
@@ -81,5 +87,5 @@ def render_graph_answer(answer: dict, context: dict) -> dict:
             lines.append(no_answer_text(context))
     if out['source_urls']:
         lines.extend(['', ('Zdroj: ' if len(out['source_urls']) == 1 else 'Zdroje: ') + ', '.join(out['source_urls'])])
-    out['answer'] = '\n'.join(lines)
+    out['answer'] = clean_answer(lines)
     return out
diff --git a/tools/answer_evaluation/verified_rag.py b/tools/answer_evaluation/verified_rag.py
index bd20de2..28ff089 100644
--- a/tools/answer_evaluation/verified_rag.py
+++ b/tools/answer_evaluation/verified_rag.py
@@ -12,7 +12,8 @@ from .answer_format import Claim
 from .context_policy import answer_context_issues, citation_check, allowed_urls, normalized, text_urls
 from .rag_evidence import descriptive_text, rag_evidence_context, quote_sections, quote_matches_passage
 from .answer_coverage import answer_coverage, coverage_notices
-from .presentation import content_lines, fact_sentences, no_answer_text, work_subject
+from .presentation import (content_lines, fact_sentences, no_answer_text, work_subject,
+                           clean_answer, evidence_subject, comparison_table)
 from .content_selection import facet_lines, evidence_units, missing_facets, fit_candidates, has_qualification
 
 
@@ -200,7 +201,7 @@ def rag_answer_issues(answer: dict, context: dict) -> list[str]:
             issues.append('rejected_supported_partial_fact')
     if answer['status'] == 'no_answer' and requirements.get('content'):
         query = normalized(context['question'])
-        overview = re.search(r'\b(?:comu(?:\s+sa)?\s+ven(?:uj|ov)\w*|o\s+com|opis\w*|popis\w*|obsah\w*|zhrn\w*)\b', query)
+        overview = re.search(r'\b(?:comu(?:\s+sa)?\s+ven(?:uj|ov)\w*|o\s+com|opis\w*|popis\w*|obsah\w*|zamer\w*|zhrn\w*)\b', query)
         # Only a broad overview is directly licensed by a documented assignment.
         # Do not reject a refusal about specific results/budgets merely because
         # some unrelated narrative was retrieved from the requested work.
@@ -260,7 +261,8 @@ def rag_answer_issues(answer: dict, context: dict) -> list[str]:
                     all(claim.get(key) is None or claim[key] == p.get(key) for key in ['work_type', 'year'])
                     for claim in claims)]
             eligible = any(c.get('person')==person for c in context.get('content_candidates', []))
-            if (not requirements.get('facets') or eligible) and not any(quote_matches_passage(quote,p)
+            required = eligible if requirements.get('facets') else bool(passages)
+            if required and not any(quote_matches_passage(quote,p)
                        for p in passages for quote in quotes):
                 issues.append('missing_bound_content_evidence')
             if requirements.get('distinguish_progress') and not requirements.get('facets'):
@@ -347,8 +349,8 @@ def comparison_lines(answer: dict, context: dict) -> list[str]:
                 key = section['kind'] if section['kind'] in roles else 'other'
                 roles[key].append({'source_id': quote['source_id'], 'quote': section['text']})
         for role, label in [('assignment', 'Zadanie a ciele'), ('progress', 'Zaznamenaný stav a výsledky')]:
-            lines.append('**' + label + ':**')
             if roles[role]:
+                lines.append('**' + label + ':**')
                 units = overview_excerpts({'evidence_quotes': roles[role]}, context)
                 if role == 'assignment':
                     lines.extend(content_lines(units, person, kind, context['question'])[1:])
@@ -359,13 +361,12 @@ def comparison_lines(answer: dict, context: dict) -> list[str]:
                              ('nie je zadanie doložené.' if role == 'assignment' else 'nie sú dosiahnuté výsledky doložené.'))
             lines.append('')
         if roles['task']:
-            lines.append('**Ďalšie doložené úlohy (plán):**')
+            lines.append('**Plánované úlohy:**')
             units = overview_excerpts({'evidence_quotes': roles['task']}, context)
             lines.extend(content_lines(units, person, kind, context['question'])[1:])
             lines.append('')
         if roles['other']:
-            lines.extend(['**Ďalšie doložené podklady:**',
-                          *overview_excerpts({'evidence_quotes': roles['other']}, context), ''])
+            lines.extend([*overview_excerpts({'evidence_quotes': roles['other']}, context), ''])
     return lines
 
 
@@ -408,7 +409,15 @@ def render_rag_answer(answer: dict, context: dict) -> str:
                     if quote_matches_passage(quote,p)), {})
                 identity = (passage.get('person'), passage.get('work_type'), passage.get('year'))
                 groups.setdefault(identity, []).append(selection['text'])
-            for (person, kind, _year), selections in groups.items():
+            table = None
+            if style == 'overview' and len(groups)>1 and re.search(r'\bporovnaj\w*\b',normalized(context.get('question',''))):
+                from .presentation import goal_text
+                table=comparison_table(['Oblasť',*(evidence_subject(person,context) for person,_,_ in groups)],
+                    [('Zameranie podľa podkladov',['\n'.join(goal_text(text)[0] for text in units)
+                                                   for units in groups.values()])])
+            if table is not None:
+                lines.extend(table)
+            for (person, kind, _year), selections in ([] if table is not None else groups.items()):
                 if style == 'overview':
                     units = selections
                 else:
@@ -426,6 +435,10 @@ def render_rag_answer(answer: dict, context: dict) -> str:
                       ': „' + quote['quote'] + '“' for quote in quotes]
         if answer.get('count') is not None:
             lines.append(f'Počet: {answer["count"]}.')
+        if req.get('content') and not req.get('facets'):
+            from .rag_evidence import missing_content_people
+            for person in missing_content_people(answer,context):
+                lines.extend(['',f'{person}: v poskytnutých podkladoch chýba opis požadovanej práce alebo projektu.'])
     elif answer['status'] == 'no_answer':
         if context.get('requirements', {}).get('facets') and context['requirements']['people']:
             lines.extend(facet_lines(answer,context))
@@ -438,7 +451,7 @@ def render_rag_answer(answer: dict, context: dict) -> str:
     lines.extend(coverage_notices(answer_coverage(answer, context)))
     if answer.get('source_urls'):
         lines.extend(['', ('Zdroj: ' if len(answer['source_urls']) == 1 else 'Zdroje: ') + ', '.join(answer['source_urls'])])
-    return '\n'.join(lines)
+    return clean_answer(lines)
 
 
 def generate_verified_rag(question: str, payload: dict, complete) -> dict: