cross-linker

The Skill discovers missing cross-references by scanning markdown pages for unlinked mentions of titles, aliases, and distinctive concepts, then inserts appropriate wikilinks. It ranks candidate links by confidence, adding only high-certainty extracted and inferred links while avoiding self-links and frontmatter.

The best way to maintain backlinks for orphaned pages is to run automated vault maintenance that scans for unlinked entity mentions and inserts appropriate wikilinks. This weaves new pages into the knowledge graph by indexing titles, aliases, and tags to establish missing connections.

Cross-referencing respects AGENTS.md structure by enforcing fixed entity subfolder rules so entities link into correct locations. The linker also refreshes entity counters and updates affinity scores as required by the vault configuration.

Yes, you can automatically connect related pages without creating self-links by using conservative link insertion logic. The process scans for unlinked mentions and adds only extracted and inferred wikilinks while explicitly avoiding self-links, code blocks, and frontmatter.

To fix a disconnected knowledge graph after large ingestion, run an automated cross-linker to discover missing cross-references and insert appropriate wikilinks. It scans newly added pages, detects unlinked mentions of existing titles or aliases, and updates log and hot-cache entries.

The main limitation of automated wiki link insertion is that it only adds extracted and inferred links with reasonable confidence, avoiding low-certainty connections. It also skips adding links inside frontmatter and code blocks to preserve formatting integrity.