mirror of
https://github.com/supermemoryai/supermemory.git
synced 2026-10-27 06:15:15 +00:00
4347970d16
Fixes v5 ingestion and document-operation docs. - **Query params return `400`.** `dreaming`, `fileType` and `mimeType` were documented as query params; they are JSON body or multipart form fields. - **Errors.** Out-of-balance quota is `402`, not `429`; `429` is only rate limiting. Added the `409` returned during a namespace merge. - **Statuses.** There is no `processing` status. The real ones are queued, extracting, chunking, embedding, done and failed; treat anything else as in progress. - **`group` on updates.** `group` is ingestion-side: it picks which existing memories serve as context while content is processed. Changing it only takes effect on re-ingestion, so it's sent with `content`/`file`. The "group-only PATCH" claims are removed from the migration doc. - **Lists, deletes and forget:** - Valid `sort` values per type, a `limit` max of 100, and `include.forgotten`. - Deletes skip documents that are still processing. - Forget-matching has a 100-match cap and a 2000-character `query` limit. - **Failed documents stay `failed`.** The docs said they're deleted after 2 minutes. They actually stay `failed`, and resending the same `id` reprocesses them. - **Smaller fixes:** - Metadata allows `string[]`. - The batch response field is `count`. - `audio` is a valid `fileType`. - Chunk overlap is 200 characters. - URL extraction no longer claims Markdowner. - Search examples no longer index the response object as an array. Checked against the current API and the published SDKs (TS 5.0.1, Python 5.0.0). Legacy v3/v4 docs are unchanged apart from factual fixes. 🤖 Generated with [Claude Code](https://claude.com/claude-code)
89 lines
3 KiB
Text
89 lines
3 KiB
Text
v5 scopes forgetting to one namespace and returns the same result envelope for exact and semantic requests.
|
||
|
||
### Choose exact or semantic forgetting
|
||
|
||
| Intent | v5 operation |
|
||
| --- | --- |
|
||
| Forget reviewed memory IDs | `DELETE /ns/{namespace}/memories` |
|
||
| Find memories by meaning | `DELETE /ns/{namespace}/memories/semantic` |
|
||
|
||
### Forget exact IDs
|
||
|
||
```bash
|
||
DELETE /ns/user_1/memories
|
||
Content-Type: application/json
|
||
|
||
{"ids":["mem_1","mem_2"]}
|
||
```
|
||
|
||
Send 1–500 IDs. The response reports successful IDs in `matches` and missing, already forgotten, or out-of-namespace IDs in `errors`, so HTTP success does not imply every requested ID changed.
|
||
|
||
### Preview a semantic request
|
||
|
||
```bash
|
||
DELETE /ns/user_1/memories/semantic
|
||
Content-Type: application/json
|
||
|
||
{"query":"outdated home address","dryRun":true}
|
||
```
|
||
|
||
`dryRun: true` performs selection without changing memory state. `dryRun: false` forgets the memories selected when that request executes. `query` is 1–2000 characters, and each request selects at most 100 memories. After applying a batch, repeat the preview-and-apply workflow to find more; repeated previews alone do not paginate and may return the same matches.
|
||
|
||
### Avoid selection drift
|
||
|
||
```text
|
||
semantic request with dryRun: true
|
||
|
|
||
v
|
||
review matches[].id
|
||
|
|
||
v
|
||
exact DELETE with reviewed IDs
|
||
```
|
||
|
||
Use this workflow when a human or policy must approve the exact set. Re-running the semantic request with `dryRun: false` can select a different set if memories changed after preview.
|
||
|
||
### Read the normalized response
|
||
|
||
```json
|
||
{
|
||
"count": 1,
|
||
"matches": [{ "id": "mem_1", "memory": "Old address" }],
|
||
"errors": [{ "id": "mem_2", "error": "Memory not found" }]
|
||
}
|
||
```
|
||
|
||
`count` always equals `matches.length`. Semantic requests always return an empty `errors` array. Both dry-run and applied semantic requests use this shape; the payload alone does not replace your record of which mode was sent.
|
||
|
||
### With the SDK
|
||
|
||
<CodeGroup>
|
||
```ts Legacy
|
||
await client.memories.forget({ id: "mem_1" })
|
||
await client.memories.forgetMatching({ query: "outdated home address" })
|
||
```
|
||
|
||
```ts v5
|
||
const preview = await supermemory.memories.forgetMatching("user_1", {
|
||
query: "outdated home address",
|
||
dryRun: true,
|
||
})
|
||
|
||
await supermemory.memories.forget("user_1", {
|
||
ids: preview.matches.map((m) => m.id),
|
||
})
|
||
```
|
||
</CodeGroup>
|
||
|
||
`dryRun` is required on `forgetMatching`. Both calls return `{ count, matches, errors }`.
|
||
|
||
### Removed memory-write routes
|
||
|
||
Direct v4 memory creation and version updates (`client.memories.add`, `client.memories.updateMemory`) have no v5 replacement. Ingest source material through document routes and update the canonical document when facts change.
|
||
|
||
### Verification
|
||
|
||
- Exercise all-success, partial-success, duplicate, unknown, and cross-namespace ID sets.
|
||
- Confirm dry runs leave matched memories recallable.
|
||
- Apply reviewed IDs exactly and confirm normal recall excludes them.
|
||
- Persist the request mode alongside audit logs for semantic operations.
|