Add knowledge graph null result gap guard

[yunrongy424-oss]

@algora-pbc /claim #17

Summary

• Add a distinct null-result and knowledge-gap signal guard for SCIBASE scientific knowledge graph recommendations.
• Check extracted graph edges before discovery-mode recommendations, graph search, or digest surfacing so null, negative, inconclusive, and failed-replication evidence is not hidden behind positive/high-citation links.
• Surface underexplored concept intersections as knowledge gaps and include redacted audit logs in reviewer packets.
• Include reviewer evidence under knowledge-graph-null-result-gap-guard/reports/: JSON, Markdown, SVG, and demo.mp4.

Non-overlap

This slice is separate from broad graph foundations, recommendation diversity, ethics provenance, temporal validity, biological accession crosswalks, multilingual aliasing, software compatibility, funder/clinical lineage, geospatial provenance, and graph conflict-arbiter submissions. It focuses specifically on preserving null-result/failed-replication signals and surfacing underexplored intersections.

Validation

• node knowledge-graph-null-result-gap-guard/test.js -> passed
• node --check knowledge-graph-null-result-gap-guard/index.js
• node --check knowledge-graph-null-result-gap-guard/demo.js
• node --check knowledge-graph-null-result-gap-guard/test.js
• python -m py_compile knowledge-graph-null-result-gap-guard/make-demo-video.py
• node knowledge-graph-null-result-gap-guard/demo.js -> regenerated reports
• python knowledge-graph-null-result-gap-guard/make-demo-video.py -> regenerated reports/demo.mp4
• MP4 first frame decoded successfully as 1280x720
• Report leakage scan found no synthetic email/token strings
• git diff --cached --check -> passed before commit

[yunrongy424-oss]

Low-frequency claim-completeness follow-up.

I do not currently see a Bounty claim label or public Algora claim page for this PR, so I wanted to confirm whether the claim was recognized correctly. The PR body includes @algora-pbc /claim #17, and the validation evidence is listed there:

• node knowledge-graph-null-result-gap-guard/test.js passed
• node --check passed for the touched JS files
• python -m py_compile passed for the demo-video script
• JSON/Markdown/SVG reports and reports/demo.mp4 were regenerated
• MP4 first-frame decode/read check passed
• report leakage scan found no synthetic email/token strings
• git diff --cached --check passed before commit

Is there any missing claim metadata, demo evidence, or validation detail needed to move this into reward/payment review?