fix(pipeline): retry on 0-entity response, save raw debug, improve template

- SourcePipeline: retry split_entities stage once when 0 entity delimiters
  are found (free-tier models intermittently return short non-formatted
  responses); save raw LLM response to <stage>-raw.md alongside prompts
- Return None (pause pipeline) rather than writing empty view file when
  no entities found after max retries
- _http.py: wrap json.JSONDecodeError in LLMAPIError with body preview
- extract-entities.md: add explicit H2-heading format example to Output
  Format section to prevent models from using inline "Section:" format

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

examples/infospace-with-history/templates/extract-entities.md

@@ -52,3 +52,35 @@ already contains it. Only extract entities that are genuinely new.
 
 Output each entity as a separate markdown document, delimited by
 `--- ENTITY: <entity-name> ---` markers.
+
+Use **H2 headings** (`##`) for each section inside the entity document.
+Do NOT use inline `Section:` format or H3 headings.
+
+Example of a correctly formatted entity:
+
+```
+--- ENTITY: division of labour ---
+
+# Division of Labour
+
+## Definition
+
+The separation of a work process into distinct tasks performed by specialised
+workers, increasing productivity through greater dexterity, saved time, and
+the invention of labour-saving machinery.
+
+## Source Chapter
+
+Book I, Chapter 1
+
+## Context
+
+The opening chapter's central argument, illustrated by Smith's pin factory
+example showing how dividing 18 operations dramatically increases output.
+
+## Economic Domain
+
+Production
+
+---
+```