We need to output JSON with fields: title, description (950-1100 words), specs (build_year, function, armament, crew, thickness, type, era), status. We have location: Unnamed, Country Norway, Coordinates 70.509872,24.0468755. Likely German WWII title: false. SEO/GEO guidance: Discoverability is weak; need to improve findability with precise local place names, nearby landmarks, military-heritage search intent wording.
We have existing description snippet but incomplete. We must produce a valid JSON with fields. We must follow strict rules: - Title: current title is "German Naval Bunker near Narvik". Is that appropriate? We need to check if we can rename title. Title must be short, 2-5 words, factual, confirmed by web results.
We have no web results provided. The user gave only location name, country, coordinates, and SEO guidance. No web search results. So we cannot confirm any specific name. The current title "German Naval Bunker near Narvik" might be inaccurate because coordinates are 70.509872,24.0468755.
Let's see where that is: That coordinate is near Narvik? Narvik is at approx 68.45, 17.9. The given coordinate is 70.51, 24.05, which is far north, near the Swedish border, near the city of Kiruna? Actually 70.5 N, 24.0 E is in Finnish Lapland? Let's check: 70.5 N, 24.0 E is near the town of Muonio?
Actually that's in Finland near the border with Sweden. But Norway extends to that longitude? Norway's western coast is around 13 E. 24 E is far east, in Russia? Let's approximate: 70.5 N, 24.0 E is in the Norwegian Sea? Actually 24E is east of Norway, near the Russian border? The coordinate might be in the interior of Norway, maybe near the town of Rovaniemi?
No that's 66N. 70.5N, 24E is near the town of Kautokeino? That's 69N. Actually 70.5 N, 24 E is in the Norwegian Finnmark region, near the town of Karasjok? Karasjok is around 69.5 N, 24E? Possibly near the village of Kautokeino (69.5N, 23E). So 70.5N, 24E is a bit north, maybe near the border with Russia?
It's in the Arctic area of Norway, near the town of Narvik? No Narvik is at 68.45N, 17.9E. So this coordinate is further north and east, maybe near the town of Hinnovd? Not sure. But it's not near Narvik. So the title "German Naval Bunker near Narvik" is likely inaccurate. We cannot claim it's near Narvik if coordinates are far away.
So we should not rename title to something else unless we have confirmation. The current title is "German Naval Bunker near Narvik". That is likely incorrect given coordinates. We should perhaps leave title unchanged? The rule: Title renaming only if current title is "Unnamed" or vague.
Here it's not vague; it's specific but possibly inaccurate. However we cannot confirm it's correct. The rule says we can only rename if we have confirmation from web results. Since we have no web results, we cannot confirm. So we should keep the title as is? But maybe we should set title to something else?
The rule says "Only suggest a new title if the current title is 'Unnamed' or extremely generic (like just 'Bunker')." The current title is not extremely generic; it's specific but possibly wrong. We should not change it without verification. So we keep "German Naval Bunker near Narvik" as title?
But maybe we should adjust to be more accurate: maybe "German Coastal Bunker in Finnmark"? But we cannot guess. So keep original title. But we must ensure title is 2-5 words. "German Naval Bunker near Narvik" is 5 words? Actually "German" (1) "Naval" (2) "Bunker" (3) "near" (4) "Narvik" (5).
That's 5 words, okay. But is it confirmed? Not sure. We have no web results to confirm. The rule: Only rename if current title is Unnamed or extremely generic. Since it's not, we should not rename. So we keep the title as is. But we must verify status: must be "verified" only if web results explicitly confirm location exists and matches coordinates.
We have no web results. So status should be "unverified". Because we cannot verify. Specs: we have no confirmed data. build_year not confirmed; function not confirmed; armament not confirmed; crew not confirmed; thickness not confirmed; type must be one of allowed list. The location is in Norway, likely a German WWII bunker?
But we cannot assume. The user gave "Likely German WWII title: false". So it's not likely a German WWII title. So we cannot set type to "Military Bunker"? Actually type must be one of allowed list. Could be "Coastal Battery", "Flak Tower", "Ammunition Storage", etc. But we need to choose based on evidence.
Since we have no evidence, maybe we choose "Other"? But "Other" is allowed. However we might want to choose "Military Bunker" if it's a generic bunker. The allowed list includes "Military Bunker". Yes it's there. So we can set type to "Military Bunker". Era: we can set "WWII" maybe?
But we need confirmation. Not given. Could be Cold War? Not sure. Since no evidence, maybe we set era to "Cold War"? But that would be speculation. Better to leave era unspecified? The spec requires era string. Must provide a string. Could we leave it as "WWII"? That might be inaccurate.
The rule: Specs: only confirmed data. If not directly supported, omit it entirely. So we should omit fields that are not confirmed. But the JSON schema expects those fields; we can include them with null? The instruction says "specs: ". It doesn't say we can omit fields; but we can include them with maybe empty string?
But better to include only confirmed data. However the spec includes fields: build_year, function, armament, crew, thickness, type, era. If we cannot confirm, perhaps we should omit them? But the JSON must be valid; we can include them with null or empty string? The instruction didn't say we can omit; but earlier says "If a spec field is not directly supported by the web results or the location name/type, omit it entirely." So we should omit any spec fields that are not confirmed.
But the JSON must still have the key "specs"? It can have subkeys? If we omit subkeys, the JSON would be missing them. But the instruction says "omit it entirely" meaning we should not include that key? But the top-level spec is an object; we can include only the fields we have data for.
So we can include only those we know: maybe "type" and "era"? But we need to confirm. We don't know type. Could we set type to "Other"? That is a allowed value, but is it confirmed? Not really. But we could set type to "Other" as a fallback if we cannot determine. The rule: type must be exactly one of the allowed list.
"Other" is allowed. So we can set type to "Other". That is a safe choice. For era, we could set "WWII" maybe? But not confirmed. Could set "Cold War"? Not confirmed. Could set "Unspecified"? Not allowed; must be one of the listed era examples? Actually era is just a string, not restricted to list.
It can be any string like "WWII". So we could set era to "WWII" but that would be speculation. Better to omit era? But the spec expects era field; we could set it to null? Not sure. The instruction: "era": "string (e.g. WWII, Cold War)". It doesn't say it must be from a list; just example.
So we can set era to "Unverified". But that might be odd. Perhaps we should omit era field entirely if not confirmed. But the spec expects it; but the instruction says "only confirmed data". So we could omit it. But the JSON must be valid; we can simply not include era key. However the spec definition includes era; but the output is just a JSON with that key?
The instruction didn't explicitly require to include all fields; just to output a valid JSON object with those fields. It says "Respond with a valid JSON object (no markdown fences) with these fields:" then lists fields including specs with subfields. So we must include all top-level fields: title, description, specs, status.
Within specs, we must include the subfields? Probably yes, but we can set them to null or empty string if not confirmed? The instruction: "If a spec field is not directly supported by the web results or the location name/type, omit it entirely." That suggests we can omit the subfield from the specs object.
So we can have specs: etc. But we need to decide which subfields to include. Given we have no confirmed data, we might only include "type": "Other"? But is that confirmed? Not really. But we can set type to "Other" as a safe default. For other fields, we can omit them. So specs could be: .
That would be a valid subfield. But we also need to include "era"? Not confirmed, so omit. So specs: . But we also need to include "function", "armament", "crew", "thickness", "build_year". Those we omit. Thus specs: . Now description: Must be 950-1100 words, English, third-person encyclopaedic tone, with inline citation links (3-8) using only source URLs present in provided web/Bing context.
But we have no web results provided. So we have no URLs to link. The rule: Inline source links required. Use 3-8 inline markdown links in the description (for places, operations, organizations, technical terms) using ONLY source URLs present in the provided web/Bing context. Since there are no source URLs, we cannot include any links.
But the rule says we must include 3-8 inline links. If we cannot, maybe we need to still produce description but without links? The rule is mandatory. However we have no source URLs; maybe we can still produce description without links? The rule says "Inline source links required." But