@ai-matrx/agents 0.21.4 → 0.21.6

package/CHANGELOG.md

@@ -1,5 +1,33 @@
 # Changelog
 
+## 0.21.6
+
+Automatic changed-only republish (docs/metadata drift since the last tag — see
+`git diff npm/matrx-agents/v0.21.5..npm/matrx-agents/v0.21.6 -- apps/shared/matrx-agents`).
+No source changes intended and no consumer action required.
+
+## 0.21.5
+
+### Fixed (in `./context`)
+
+- A record that merely has a `content` key (a note `{ id, title, content }`) was treated as a rich
+  envelope: its size was measured on `.content` only, and it was sent as
+  `{ id, title, content, label, max_inline_chars }`, which the server reads as raw content — the
+  page limit was silently lost. Envelope detection now matches the server exactly: own `content`
+  and every key in `RICH_ENVELOPE_KEYS` (aidream `_RICH_FORM_KEYS`); anything else is wrapped as
+  `{ content: record, … }`. Applies to `resolveContextRow` sizing, `buildContextWire` and
+  `serverDefaultMax`.
+
+### Added
+
+- `RICH_ENVELOPE_KEYS`, `isRichEnvelope(value)`, `splitContextValue(value)` (the twin of aidream
+  `_split_rich_context_value`). Corpus `envelope_cases`, run here and by aidream pytest.
+
+### Consumer action
+
+- matrx-frontend: no API change; drop any local "is this an envelope" check in favor of
+  `isRichEnvelope` / `splitContextValue` from `@ai-matrx/agents/context`.
+
 ## 0.21.4
 
 Automatic changed-only republish (docs/metadata drift since the last tag — see