ur-agent 1.65.11 → 1.65.13
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 +36 -0
- package/dist/cli.js +1118 -493
- package/docs/VALIDATION.md +1 -1
- package/documentation/index.html +1 -1
- package/extensions/jetbrains-ur/build.gradle.kts +1 -1
- package/extensions/vscode-ur-inline-diffs/package.json +1 -1
- package/package.json +1 -1
- package/technical/04-tools.md +52 -8
- package/technical/05-providers-and-models.md +9 -6
- package/technical/09-multi-agent.md +12 -4
- package/technical/README.md +1 -1
package/CHANGELOG.md
CHANGED
|
@@ -1,5 +1,41 @@
|
|
|
1
1
|
# Changelog
|
|
2
2
|
|
|
3
|
+
## 1.65.13
|
|
4
|
+
|
|
5
|
+
- Recovered otherwise valid `AskUserQuestion` calls whose UI header exceeds
|
|
6
|
+
the 12-character chip width. Native structured calls, bare/wrapped JSON, and
|
|
7
|
+
explicit-choice recovery now compact only that bounded presentation field;
|
|
8
|
+
questions, choices, labels, descriptions, previews, metadata, and selection
|
|
9
|
+
behavior remain unchanged, while unsafe or grossly oversized headers still
|
|
10
|
+
fail validation.
|
|
11
|
+
- Allowed strictly parsed syntax verification after a task-free one-shot
|
|
12
|
+
Write. `node --check <file>` and the bounded transcript-produced HTML script
|
|
13
|
+
checker bypass only the task-list requirement; arbitrary Node evaluation,
|
|
14
|
+
extra commands, redirects, backgrounding, sandbox overrides, and
|
|
15
|
+
permission-time rewrites fail closed, and normal Bash permission/sandbox
|
|
16
|
+
checks still apply.
|
|
17
|
+
|
|
18
|
+
## 1.65.12
|
|
19
|
+
|
|
20
|
+
- Forced explicit weak-model decision menus through the real
|
|
21
|
+
`AskUserQuestion` UI. A provider-neutral end-turn guard accepts only one
|
|
22
|
+
canonical Ask object from explicit reasoning or one complete rigid Markdown
|
|
23
|
+
choice menu, validates it unchanged against the live tool schema, and rejects
|
|
24
|
+
examples, ambiguity, ordinary prose, workers, and headless sessions.
|
|
25
|
+
- Removed two task-list dead ends without weakening the mutation boundary.
|
|
26
|
+
Live plan mode may run only the exact current plan-directory bootstrap Bash
|
|
27
|
+
shape before tasks exist, and the last actionable task remains `in_progress`
|
|
28
|
+
when a file mutation has no later successful observable check. The same task
|
|
29
|
+
can then be verified or corrected instead of creating a duplicate.
|
|
30
|
+
- Improved exact `Edit` mismatch recovery for stale HTML blocks. Diagnostics
|
|
31
|
+
now select the most distinctive verified line near the real target, such as
|
|
32
|
+
`<script>`, rather than an unrelated generic closing tag; edits remain
|
|
33
|
+
fail-closed and never apply fuzzy replacements.
|
|
34
|
+
- Raised the default Ollama Cloud timeout for the known-slow Kimi K2.7 family
|
|
35
|
+
from 120 to 300 seconds in both streaming and permitted non-streaming paths.
|
|
36
|
+
Explicit timeouts and the stricter remote-session ceiling still take
|
|
37
|
+
precedence.
|
|
38
|
+
|
|
3
39
|
## 1.65.11
|
|
4
40
|
|
|
5
41
|
- Hardened `AskUserQuestion` for weaker models without inventing user intent.
|