ur-agent 1.85.0 → 1.85.1
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 +21 -0
- package/README.md +22 -1
- package/dist/cli.js +399 -114
- package/docs/CONFIGURATION.md +22 -0
- package/docs/TROUBLESHOOTING.md +29 -0
- package/docs/USAGE.md +18 -1
- package/docs/VALIDATION.md +13 -1
- package/docs/providers.md +21 -4
- 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/CHANGELOG.md
CHANGED
|
@@ -1,5 +1,26 @@
|
|
|
1
1
|
# Changelog
|
|
2
2
|
|
|
3
|
+
## 1.85.1
|
|
4
|
+
|
|
5
|
+
- Added complete Anthropic workspace routing for identity-linked Claude API
|
|
6
|
+
keys. A validated workspace ID can be configured with
|
|
7
|
+
`anthropic.workspace_id`, `ANTHROPIC_WORKSPACE_ID`, or
|
|
8
|
+
`ur connect anthropic-api --workspace-id`; UR applies the matching header to
|
|
9
|
+
model discovery, diagnostics, streaming and non-streaming Messages requests,
|
|
10
|
+
and token counting while leaving workspace-scoped keys unchanged.
|
|
11
|
+
- Preserved Anthropic's real missing-workspace response and replaced the
|
|
12
|
+
misleading provider/model compatibility failure with an actionable setup
|
|
13
|
+
instruction. Live model caches are partitioned by workspace so one workspace
|
|
14
|
+
cannot reuse another workspace's catalog.
|
|
15
|
+
- Made OpenAI permanent account failures fail immediately instead of looking
|
|
16
|
+
like slow inference. Streaming HTTP error bodies are decoded safely, exact
|
|
17
|
+
provider error codes such as `billing_not_active` and `insufficient_quota`
|
|
18
|
+
are treated as non-retryable, and genuine transient rate limits continue to
|
|
19
|
+
use retry and backoff behavior.
|
|
20
|
+
- Isolated provider-discovery tests from live keychain credentials and updated
|
|
21
|
+
public, configuration, troubleshooting, validation, provider, and technical
|
|
22
|
+
documentation for both behaviors.
|
|
23
|
+
|
|
3
24
|
## 1.85.0
|
|
4
25
|
|
|
5
26
|
- Split NVIDIA Build into two truthful provider surfaces backed by one saved
|
package/README.md
CHANGED
|
@@ -299,6 +299,7 @@ ur config set provider.fallback ollama
|
|
|
299
299
|
ur config set openrouter.routing auto
|
|
300
300
|
ur config set openrouter.preferred_min_throughput 40
|
|
301
301
|
ur config set openrouter.preferred_max_latency 3
|
|
302
|
+
ur config set anthropic.workspace_id wrkspc_...
|
|
302
303
|
```
|
|
303
304
|
|
|
304
305
|
`provider.fallback` records a recovery provider for `ur provider doctor`
|
|
@@ -332,6 +333,11 @@ store records identifiers and status only. Compacted context is persisted only
|
|
|
332
333
|
when `UR_OPENAI_RESPONSES_STATE_KEY` contains a 32-byte encryption key. Return
|
|
333
334
|
to the default with `ur config set openai_transport chat-completions`.
|
|
334
335
|
|
|
336
|
+
Both OpenAI transports retry genuine temporary rate limits. Permanent
|
|
337
|
+
machine-coded account or billing responses such as `billing_not_active` and
|
|
338
|
+
`insufficient_quota` are reported immediately, so an inactive project does not
|
|
339
|
+
look like minute-long model latency.
|
|
340
|
+
|
|
335
341
|
OpenRouter `auto` routing leaves tool turns to Auto Exacto and optimizes
|
|
336
342
|
non-tool turns for end-to-end throughput. The routing, fallback,
|
|
337
343
|
strict-parameter, rolling performance preference, service-tier, and supported
|
|
@@ -351,6 +357,21 @@ ur config set openrouter.speed fast # standard | fast
|
|
|
351
357
|
Direct Anthropic requests automatically preserve UR's prompt-cache
|
|
352
358
|
breakpoints and enable per-tool fine-grained input streaming, reducing repeat
|
|
353
359
|
prefill work and exposing large tool arguments as Claude generates them.
|
|
360
|
+
Anthropic identity-linked keys that can act across workspaces also require a
|
|
361
|
+
workspace on discovery, inference, streaming, and token-count calls. Configure
|
|
362
|
+
the Console workspace ID once (it is not a secret):
|
|
363
|
+
|
|
364
|
+
```sh
|
|
365
|
+
ur config set anthropic.workspace_id wrkspc_...
|
|
366
|
+
# or for one environment:
|
|
367
|
+
export ANTHROPIC_WORKSPACE_ID=wrkspc_...
|
|
368
|
+
```
|
|
369
|
+
|
|
370
|
+
UR sends `anthropic-workspace-id` only when configured. Workspace-scoped keys
|
|
371
|
+
continue to work without it. Use `auto` to clear the saved selection. If the
|
|
372
|
+
key and workspace are being connected together, `ur connect anthropic-api
|
|
373
|
+
--workspace-id wrkspc_...` accepts the same value.
|
|
374
|
+
|
|
354
375
|
Anthropic's premium research-preview fast tier is opt-in and is sent only for
|
|
355
376
|
Claude Opus 5 or Opus 4.8:
|
|
356
377
|
|
|
@@ -380,7 +401,7 @@ ur connect logout openai-api # clear a stored key
|
|
|
380
401
|
| Provider | Access type | Runtime kind | Legal path |
|
|
381
402
|
| --- | --- | --- | --- |
|
|
382
403
|
| OpenAI API | API key | UR-native | `OPENAI_API_KEY` or `ur connect openai-api` |
|
|
383
|
-
| Claude API | API key | UR-native | `ANTHROPIC_API_KEY` or `ur connect anthropic-api` |
|
|
404
|
+
| Claude API | API key | UR-native | `ANTHROPIC_API_KEY` or `ur connect anthropic-api`; identity-linked keys also use `ANTHROPIC_WORKSPACE_ID` or `anthropic.workspace_id` |
|
|
384
405
|
| Gemini API | API key | UR-native | `GEMINI_API_KEY` or `ur connect gemini-api` |
|
|
385
406
|
| OpenRouter | API/router | UR-native | `OPENROUTER_API_KEY` or `ur connect openrouter` |
|
|
386
407
|
| NVIDIA Agentic | hosted/server API | UR-native | `NVIDIA_API_KEY` or `ur connect nvidia-nim`; configurable `base_url` |
|