Lidang Public Ideas

This hub tracks Lidang’s public idea stream as a high-frequency corpus. The system favors batch summaries, confidence labels, and topic clustering over one page per short post.

Current Corpus Shape

• x-profile-probe: 1
• x-status-manual-search: 1
• youtube-video: 15

Core Use

• Use lidang-idea-taxonomy to navigate topics.
• Use 2026-05-16-what-indexes-did-lidang-recommend for the SP500 / Nasdaq-100 clarification.
• Use lidang-weekly-digests for chronological batches.
• Use high-signal single-item pages only when an item is substantial enough to justify durable retrieval.
• Use source confidence labels before citing any X-derived item.

Maintenance Rules For Future Agents

• Treat HTML as a first-class raw source format for X/profile probes; store reachable snapshots under raw/lidang-public/html/.
• Use semantic_hash rather than raw full HTML hash for wiki churn decisions.
• Never create a second item for a stable ID that already exists in raw/lidang-public/manifest.json.
• Continue backfill attempts for older reachable YouTube/X items, but add them only when their stable IDs are absent.
• Do not flood wiki/index.md with every short post.
• Keep X mirror text out of public pages unless a license-safe source exists.
• Record failed mirror probes because failures are operationally useful.
• Prefer weekly synthesis over raw accumulation.

Related

• lidang
• 2026-05-16-lidang-public-corpus
• lidang-idea-taxonomy
• lidang-weekly-digests

Counterpoints And Gaps

• The current corpus is not a complete X archive.
• YouTube RSS only exposes recent videos, so historical video backfill remains dependent on optional tooling or future source access.
• Short-form public statements can be context-dependent and should not be over-interpreted without source review.