docs(mcp): simplify document status section

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
This commit is contained in:
MaheshtheDev 2026-10-27 00:35:07 +00:00
parent 4d9c13c2a4
commit 62717be140

View file

@ -58,13 +58,18 @@ The client discovers the OAuth authorization server through
| `add_memory` | Accept a document for asynchronous memory extraction (legacy `action: "forget"` remains supported) |
| `forget_memory` | Forget a memory matching the supplied content |
`add_memory` success confirms document acceptance, not extracted or searchable
memories. In `list_documents` and `get_document`, `status: done` means content
processing completed. The separate `dreamingStatus`, when returned by the API,
is `dreaming` while generation is pending, running, or awaiting recovery after
failure, and `done` when generation settled or no memories were needed. Neither
field guarantees a nonzero memory count or that memory-vector writes finished.
These tools pass through the existing API signal without additional requests.
### Document status
`list_documents` and `get_document` return two status fields:
| Field | Value | Meaning |
| --- | --- | --- |
| `status` | `done` | Content is processed and readable |
| `dreamingStatus` | `dreaming` | Memory extraction is pending, running, or retrying |
| `dreamingStatus` | `done` | Memory extraction settled, possibly with no memories |
`dreamingStatus` appears only when the API returns it. Neither `done` means
memories are searchable yet.
### MCP App launchers