- Add outputs/prompts/article-summary-prompt.txt with deep knowledge
note schema (core_conclusion, main_argument, key_methods, etc.)
- Add ArticleSummaryResult Pydantic model for the new schema
- Add validate_article_summary_payload() validator
- Add optional validator param to run_loop_payload(); defaults to
existing validate_llm_result_payload so daily digest is unaffected
- Update article_summary.py: new prompt path, new validator, new
Markdown output with Chinese section headers
Extract three internal helpers to reduce the main function from ~330
lines to ~80 lines:
- _process_item(): handles per-item extract/summarize/filter/build
- _build_and_persist_delivery(): builds payload and persists keyword index
- _build_run_report(): assembles status counts and run report dict
No behavior changes. External signature and return shape unchanged.
Add explicit authoring constraint in filter-rule-engine-usage.md:
all drop rules must set stop_on_match=true to prevent low-priority
drop rules from overriding high-priority keep decisions.
Eliminate the NamedTemporaryFile workaround in server.py by adding a
`context` dict parameter to run_freshrss_pipeline. The pipeline resolves
filter context from the dict first, falling back to context_path, then
defaulting to an empty FilterContext. Remove unused json/tempfile imports.
Added explanatory comments to server.py, freshrss_pipeline.py, pipeline.py,
freshrss.py, keyword_index.py, summary_loop.py, and filters/engine.py.
Also added CODE_REVIEW.md with prioritized improvement suggestions.
wdm