snipara-companion 3.8.0 → 3.10.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.md +11 -0
- package/dist/index.d.ts +3099 -3088
- package/dist/index.js +820 -315
- package/docs/MEMORY_PILOT.md +48 -0
- package/docs/TASK_JOURNEY.md +31 -3
- package/package.json +1 -1
package/CHANGELOG.md
CHANGED
|
@@ -2,6 +2,17 @@
|
|
|
2
2
|
|
|
3
3
|
Release notes for `snipara-companion`, newest first.
|
|
4
4
|
|
|
5
|
+
## 3.10.0 (unpublished)
|
|
6
|
+
|
|
7
|
+
- Add opt-in `agent-context memory-pilot` for automatic exact-source observation
|
|
8
|
+
validation/invalidation and fresh code/review/resume receipts. Recoverable
|
|
9
|
+
evidence failures stay candidates; semantic judgments escalate only when the
|
|
10
|
+
deterministic verifier cannot decide safely. No hosted memory policy changes.
|
|
11
|
+
|
|
12
|
+
## 3.9.0
|
|
13
|
+
|
|
14
|
+
- Project routine task supervision into public task cards, identifying the agent or human responsible for the next bounded action. Keep proof, execution and integration authority explicit.
|
|
15
|
+
|
|
5
16
|
## 3.8.0 — candidate
|
|
6
17
|
|
|
7
18
|
- Add standalone Sandbox validation adapters with explicit scoped environment choices, preserved failure evidence and no silent host fallback.
|