Fix grounded assistant cross-domain responses

CHANGELOG.md

@@ -7,6 +7,17 @@
 
 # Changelog
 
+## Sprint 215 Grounded assistant cross-domain reliability (2026-07-16)
+
+- Raised the bounded local Ollama answer budget from 700 to 1,200 tokens after
+  a live Mol profile exhausted the smaller limit; `done_reason=length` remains
+  a hard error and incomplete text is still never returned.
+- Required listed themes to be covered without adding unrelated themes and
+  kept cross-domain answers compact without weakening source, unit, estimate,
+  chronology or unsupported-metric rules.
+- Exposed the output limit in the Unraid DockerMan template and aligned all
+  Compose, runtime, example and operator documentation defaults.
+
 ## Sprint 213-214 Cross-domain area profile (2026-07-16)
 
 - Implemented one allowlisted MercatorNet WCS registry for official Flemish

backend/README.md

@@ -1114,7 +1114,7 @@ OLLAMA_ENABLED=true
 OLLAMA_BASE_URL=http://host.docker.internal:11434
 OLLAMA_DEFAULT_MODEL=qwen3.5:9b
 OLLAMA_TIMEOUT_SECONDS=120
-OLLAMA_MAX_OUTPUT_TOKENS=700
+OLLAMA_MAX_OUTPUT_TOKENS=1200
 OLLAMA_CONTEXT_TOKENS=16384
 ```
 
@@ -1125,7 +1125,9 @@ models with `GET /api/v1/assistant/models` and ask a grounded question through
 present in Ollama `/api/tags`. Missing water depth/bathymetry remains explicit;
 the assistant cannot turn 2D water geometry into volume. GeoIntel rejects an
 answer when Ollama reports `done_reason=length`, so a visibly truncated sentence
-is never presented as a complete result.
+is never presented as a complete result. The 1,200-token default leaves enough
+room for a compact cross-domain profile while the system prompt requires every
+explicitly requested theme and excludes unrelated themes.
 
 ## Agricultural-use parcel history
 

backend/app/core/config.py

@@ -82,7 +82,7 @@ class Settings(BaseSettings):
     ollama_base_url: str = Field(default="http://127.0.0.1:11434", validation_alias="OLLAMA_BASE_URL")
     ollama_default_model: str = Field(default="qwen3.5:9b", validation_alias="OLLAMA_DEFAULT_MODEL")
     ollama_timeout_seconds: int = Field(default=120, ge=5, le=600, validation_alias="OLLAMA_TIMEOUT_SECONDS")
-    ollama_max_output_tokens: int = Field(default=700, ge=100, le=4_000, validation_alias="OLLAMA_MAX_OUTPUT_TOKENS")
+    ollama_max_output_tokens: int = Field(default=1_200, ge=100, le=4_000, validation_alias="OLLAMA_MAX_OUTPUT_TOKENS")
     ollama_context_tokens: int = Field(default=16_384, ge=4_096, le=131_072, validation_alias="OLLAMA_CONTEXT_TOKENS")
     cors_origins: list[str] | str = Field(
         default=["http://localhost:5173", "http://127.0.0.1:5173"],

backend/app/services/geo_assistant_service.py

@@ -525,6 +525,8 @@ class GeoAssistantService:
             "Noem bij cijfers de bron en eenheid. Maak duidelijk onderscheid tussen exacte metingen en schattingen. "
             "Als is_estimate true is, noem de waarde verplicht een schatting en nooit exact. "
             "Objectaantallen zijn ondersteunend; geef betekenisvolle oppervlakte-, lengte- of bevolkingsmetriek voorrang. "
+            "Wanneer de gebruiker meerdere thema's opsomt, behandel elk gevraagd thema en voeg geen ongevraagd thema toe. "
+            "Houd het antwoord beknopt: groepeer de kernmetrieken per gevraagd thema en herhaal geen beperkingen. "
             "Beschrijf alleen waargenomen verschillen; verzin geen oorzaak, voorspelling, verzadiging of andere verklaring. "
             "Neem waarden en jaren letterlijk over en bereken zelf geen gemiddelde, tempo, oorzaak of afgeleide trend. "
             "Gebruik platte tekst met korte alinea's en opsommingen, zonder Markdown-symbolen. "

backend/tests/test_sprint202_temporal_metrics_and_ollama.py

@@ -202,12 +202,15 @@ def test_geo_assistant_sends_grounded_context_without_thinking_trace(monkeypatch
     assert captured["payload"]["think"] is False
     assert captured["payload"]["options"]["temperature"] == 0.0
     assert captured["payload"]["options"]["num_ctx"] == 16_384
+    assert captured["payload"]["options"]["num_predict"] == 1_200
     assert "Gebruik uitsluitend feiten en cijfers uit CONTEXT_JSON" in captured["payload"]["messages"][0]["content"]
     assert "scope.label is het exact geanalyseerde gebied" in captured["payload"]["messages"][0]["content"]
     assert "noem de waarde verplicht een schatting" in captured["payload"]["messages"][0]["content"]
     assert "verzin geen oorzaak, voorspelling, verzadiging" in captured["payload"]["messages"][0]["content"]
     assert "bereken zelf geen gemiddelde, tempo, oorzaak of afgeleide trend" in captured["payload"]["messages"][0]["content"]
     assert "zonder Markdown-symbolen" in captured["payload"]["messages"][0]["content"]
+    assert "behandel elk gevraagd thema en voeg geen ongevraagd thema toe" in captured["payload"]["messages"][0]["content"]
+    assert "Houd het antwoord beknopt" in captured["payload"]["messages"][0]["content"]
     assert "water_volume_available" in captured["payload"]["messages"][0]["content"]
 
 
@@ -254,6 +257,9 @@ def test_unraid_ollama_context_window_is_configurable() -> None:
     assert "OLLAMA_CONTEXT_TOKENS: ${OLLAMA_CONTEXT_TOKENS:-16384}" in compose
     assert 'Target="OLLAMA_CONTEXT_TOKENS"' in template
     assert "OLLAMA_CONTEXT_TOKENS=16384" in env_example
+    assert "OLLAMA_MAX_OUTPUT_TOKENS: ${OLLAMA_MAX_OUTPUT_TOKENS:-1200}" in compose
+    assert 'Target="OLLAMA_MAX_OUTPUT_TOKENS"' in template
+    assert "OLLAMA_MAX_OUTPUT_TOKENS=1200" in env_example
 
 
 def test_temporal_comparison_preserves_all_compatible_semantic_metrics() -> None:

deploy/unraid/README.md

@@ -200,7 +200,7 @@ OLLAMA_ENABLED=true
 OLLAMA_BASE_URL=http://host.docker.internal:11434
 OLLAMA_DEFAULT_MODEL=qwen3.5:9b
 OLLAMA_TIMEOUT_SECONDS=120
-OLLAMA_MAX_OUTPUT_TOKENS=700
+OLLAMA_MAX_OUTPUT_TOKENS=1200
 OLLAMA_CONTEXT_TOKENS=16384
 ```
 

deploy/unraid/geointel-unraid-template.xml

@@ -51,5 +51,6 @@
   <Config Name="Ollama Base URL" Target="OLLAMA_BASE_URL" Default="http://host.docker.internal:11434" Mode="" Description="Ollama API reachable from the container. The deployment maps host.docker.internal to the Unraid host gateway." Type="Variable" Display="always" Required="true" Mask="false">http://host.docker.internal:11434</Config>
   <Config Name="Default Ollama Model" Target="OLLAMA_DEFAULT_MODEL" Default="qwen3.5:9b" Mode="" Description="Preferred locally installed Ollama model. Users can select another installed model in GeoIntel." Type="Variable" Display="always" Required="true" Mask="false">qwen3.5:9b</Config>
   <Config Name="Ollama Timeout Seconds" Target="OLLAMA_TIMEOUT_SECONDS" Default="120" Mode="" Description="Maximum wait for one local assistant response." Type="Variable" Display="advanced" Required="true" Mask="false">120</Config>
+  <Config Name="Ollama Maximum Output Tokens" Target="OLLAMA_MAX_OUTPUT_TOKENS" Default="1200" Mode="" Description="Maximum generated tokens for one grounded answer. Length-truncated responses remain rejected." Type="Variable" Display="advanced" Required="true" Mask="false">1200</Config>
   <Config Name="Ollama Context Tokens" Target="OLLAMA_CONTEXT_TOKENS" Default="16384" Mode="" Description="Context window reserved for grounded GIS measurements and the generated answer." Type="Variable" Display="advanced" Required="true" Mask="false">16384</Config>
 </Container>

deploy/unraid/geointel.env.example

@@ -71,5 +71,5 @@ OLLAMA_ENABLED=true
 OLLAMA_BASE_URL=http://host.docker.internal:11434
 OLLAMA_DEFAULT_MODEL=qwen3.5:9b
 OLLAMA_TIMEOUT_SECONDS=120
-OLLAMA_MAX_OUTPUT_TOKENS=700
+OLLAMA_MAX_OUTPUT_TOKENS=1200
 OLLAMA_CONTEXT_TOKENS=16384

deploy/unraid/run-dockerman-container.sh

@@ -67,7 +67,7 @@ OLLAMA_ENABLED="${OLLAMA_ENABLED:-true}"
 OLLAMA_BASE_URL="${OLLAMA_BASE_URL:-http://host.docker.internal:11434}"
 OLLAMA_DEFAULT_MODEL="${OLLAMA_DEFAULT_MODEL:-qwen3.5:9b}"
 OLLAMA_TIMEOUT_SECONDS="${OLLAMA_TIMEOUT_SECONDS:-120}"
-OLLAMA_MAX_OUTPUT_TOKENS="${OLLAMA_MAX_OUTPUT_TOKENS:-700}"
+OLLAMA_MAX_OUTPUT_TOKENS="${OLLAMA_MAX_OUTPUT_TOKENS:-1200}"
 OLLAMA_CONTEXT_TOKENS="${OLLAMA_CONTEXT_TOKENS:-16384}"
 
 install_dockerman_metadata() {

docker-compose.unraid.yml

@@ -65,7 +65,7 @@ services:
       OLLAMA_BASE_URL: ${OLLAMA_BASE_URL:-http://host.docker.internal:11434}
       OLLAMA_DEFAULT_MODEL: ${OLLAMA_DEFAULT_MODEL:-qwen3.5:9b}
       OLLAMA_TIMEOUT_SECONDS: ${OLLAMA_TIMEOUT_SECONDS:-120}
-      OLLAMA_MAX_OUTPUT_TOKENS: ${OLLAMA_MAX_OUTPUT_TOKENS:-700}
+      OLLAMA_MAX_OUTPUT_TOKENS: ${OLLAMA_MAX_OUTPUT_TOKENS:-1200}
       OLLAMA_CONTEXT_TOKENS: ${OLLAMA_CONTEXT_TOKENS:-16384}
     ports:
       - "${GEOINTEL_FRONTEND_PORT:-1202}:80"

docker-compose.yml

@@ -70,7 +70,7 @@ services:
       OLLAMA_BASE_URL: ${OLLAMA_BASE_URL:-http://host.docker.internal:11434}
       OLLAMA_DEFAULT_MODEL: ${OLLAMA_DEFAULT_MODEL:-qwen3.5:9b}
       OLLAMA_TIMEOUT_SECONDS: ${OLLAMA_TIMEOUT_SECONDS:-120}
-      OLLAMA_MAX_OUTPUT_TOKENS: ${OLLAMA_MAX_OUTPUT_TOKENS:-700}
+      OLLAMA_MAX_OUTPUT_TOKENS: ${OLLAMA_MAX_OUTPUT_TOKENS:-1200}
       OLLAMA_CONTEXT_TOKENS: ${OLLAMA_CONTEXT_TOKENS:-16384}
     ports:
       - "${GEOINTEL_BACKEND_PORT:-8000}:8000"

docs/CODEX_EXECUTION_LOG.md

@@ -1,3 +1,29 @@
+## Sprint 215 Grounded assistant cross-domain reliability (2026-07-16)
+
+Changed:
+- Measured the exact live Mol assistant context at 12 current themes, 36
+  semantic metrics, six temporal series and 17,302 compact JSON characters.
+- Increased the bounded output allowance to 1,200 tokens and instructed the
+  local model to cover every explicitly listed theme, omit unrelated themes
+  and avoid repeated limitations.
+- Kept `think=false`, deterministic temperature, the 16,384-token context and
+  fail-closed rejection of `done_reason=length` unchanged.
+- Added the output limit to the editable Unraid template and aligned all
+  deployment defaults and documentation.
+
+Validation evidence:
+- A read-only live production-chain probe with the 1,200-token setting
+  completed without truncation against the full persisted Mol context.
+- The focused assistant suite passed 14 tests.
+- The complete readiness gate passed 713 backend tests, backend compilation,
+  105 documented API routes with three explicit binary/non-envelope routes,
+  one Alembic head and the frontend TypeScript and production build.
+- Tower deployment and browser verification follow in this pass.
+
+Next:
+- Deploy, rerun the exact six-theme Dutch question and verify the end-user
+  response and responsive assistant UI in the browser.
+
 ## Sprint 213-214 Cross-domain thematic rasters and DOV soil map (2026-07-16)
 
 Changed: