sync: harden standalone mcp and chunk recovery

This commit is contained in:
“BeeRad”
2026-04-16 09:53:56 +10:00
parent 097c4c61bb
commit c2f880d957
15 changed files with 222 additions and 279 deletions
@@ -21,7 +21,7 @@ description: "Use for graph read, write, connect, classify, or traverse operatio
- `title`: clear and specific.
- `description`: concrete object-level description, not vague summaries.
- `source`: full verbatim or canonical content of the node (transcript, article text, book passage, user's thoughts). This is what gets chunked and embedded for semantic search.
- `source`: full verbatim or canonical content of the node (transcript, article text, book passage, user's thoughts). The standalone MCP server stores this on the node. The RA-H app later chunks and embeds it for semantic search.
- For idea capture from chat, the `source` should usually be the raw user thought, not a compressed assistant summary.
- `link`: external source URL only.
- Normal writes should omit context entirely unless the user explicitly wants one.