Scan markdown files for unlinked mentions of a wiki entry's slug or title.
npx skills add https://github.com/Walliiee/agent-harness --skill memory-graph-link
Please help me install this Agent Skill. Skill: memory-graph-link Source: https://github.com/Walliiee/agent-harness/tree/main/skills/memory-graph-link Command: npx skills add https://github.com/Walliiee/agent-harness --skill memory-graph-link
This Skill helps you catch wiki entries that already mention a concept in prose but never added the [[slug]] link needed for a healthy knowledge graph. It reduces manual backlink hunting after new edits and keeps retrieval paths consistent across workspaces.
Ask the agent to scan the new wiki entry and report existing markdown files that mention its slug or title without already linking to it.
High-intent search queries and answers about installing and using this skill.
To find missing wiki backlinks, scan other markdown files across your workspaces for lexical mentions of the new page's slug or title. This identifies notes that mention the concept in prose but lack the proper [[slug]] link, keeping your knowledge graph healthy.
The best way to audit a knowledge graph for unlinked markdown mentions is to perform a lexical match for entry titles or slugs across all files. This process excludes already-linked files and index pages, returning only valid backlink candidates.
Yes, you can scan multiple workspaces for wiki backlink candidates. The process traverses every configured workspace wiki to find link opportunities wherever they appear, ensuring consistent retrieval paths across your entire knowledge base.
During a workspace scan, already-linked files and index pages are automatically excluded to control noise. The scan also applies per-file suggestion limits, ensuring you only receive relevant backlink candidates without duplicate or excessive entries.
You should run a backlink candidate scan for high-traffic wiki entries during periodic graph audits or immediately after publishing and editing knowledge-base pages. This catches older notes that mention the slug or title but lack proper linking.
michaeldyryndaMigrate legacy .ai plan layouts to slug-based directories with non-destructive moves.
terrylicaValidate markdown links for portability across Claude Code skill deployments.
ShaheerKhawajaRetrieve wiki vault pages via progressive loading and wikilink traversal for cross-domain queries.
Awannaphasch2016Perform semantic vector search against AgentDB stores using embedding similarity.
