Commit Graph
101 Commits
Author SHA1 Message Date
Magolor 5249d9a4ce fix: prefer saved credentials and use environment keys in headless (#5483)
* ui: show locked API key environment reference

* fix(ui-settings-models): skip unused pi-ai credential reference

* fix(ui-settings-models): show active provider auth source

* test(llm): cover provider auth check failures

* docs(ui-settings-models): clarify active provider authentication

* test(web): update model key copy snapshots

* fix(ui-settings-models): clarify native authentication source hint

* fix: address PR 5483 review feedback

* docs(llm): state empty auth lookup behavior

* fix: refresh auth hints with credential snapshots

* fix: require credential refresh revision

* test: cover auth status projection

* fix: refresh catalog and browser expectations after rebase

* fix: keep auth status projection internal and explicit

* fix: preserve credential source checks with signed-in account rows

* test: pin Models authentication instead of inheriting host keys

* fix: prioritize stored credentials across providers

* fix: use one environment fallback hint

* docs: refresh credential contract

* docs: refresh config catalog

* test: align provider hint verification

* fix: use provider-native auth hints and endpoints

* fix: cover oauth-capable provider auth hints

* fix: show canonical endpoint defaults

* fix: avoid misleading endpoint fallback

* fix: label catalog endpoint defaults

* fix: shorten auth hints with hover details

* fix: clarify native auth resolution

* test: align DeepSeek endpoint hint

* fix: pair help button corner shape

* fix: address credential precedence review

Record the stored-credential precedence in a superseding Agent Note and
an upgrade guide, and align the credentials subsystem page, package
READMEs, and JSDoc with the order file > launch env > $cwd/.env >
$DSH_HOME/.env.

Report each pi-ai route's catalog endpoint and native-auth environment
names through the provider directory, check the names against the
installed pi-ai resolvers, and remove the llm/checkAuth and credential
describe operations that no Models page code calls.

Restore the route-derived key reference and the home expansion that
reads the home directory only for a leading ~.

* fix: clarify model authentication guidance

* fix: explain native provider OAuth requirements

* fix: keep authentication guidance method neutral

* fix: show credential reference source in model editor

* fix: use environment authentication in headless profile

* fix: validate headless authentication and current provider metadata

* test: retain authentication guidance coverage with account catalog updates

* test: retain writable Models credentials with the pinned environment
2026-10-13 10:10:06 +08:00
Turtle 0155017b73 fix(plan-mode): exit plan mode only on approval in plan guidance
A dismissed review now returns a successful exit_plan_mode result, so
the plan policy must not treat any success as leaving plan mode. Also
update the user-facing description of a dismissed review.
2026-10-09 14:58:20 +08:00
Tianyi Cui b931f1f9f7 Merge pull request #5713 from deepseek-harness/worktree/session-title-reasoning-20261005
fix(llm,session-title): configure prepared calls from model capabilities
2026-10-07 18:02:41 +08:00
Tianyi Cui b48384223a refactor(experimental)!: classify optional packages and move prototypes 2026-10-06 16:41:04 +08:00
Tianyi Cui 3562e14f28 fix(session-title): widen the shipped title output cap to a runaway ceiling
The prompt asks for a short title, so the cap only stops runaway output:
reasoning a model cannot switch off also counts as output, and 256 tokens
truncated titles on the HFAI gateway. The measured worst title response
was 153 output tokens, so 4096 is about 25 times the observed maximum.
2026-10-06 15:12:55 +08:00
Tianyi Cui fabd805bfe fix(session-title): leave reasoning headroom in the shipped title output cap
The base bundle capped title generation at 64 output tokens, sized for a
route that disables thinking. Models whose reasoning cannot be switched
off spend output on it even at their lowest level: on the HFAI gateway,
GPT-6 Sol and GPT-6 Astra used up to 87 and 121 output tokens for
five-word titles at reasoning low, so Sol still finished with max-tokens
on most first prompts. Raise the shipped cap to 256; routes that disable
thinking stop after the title and are unaffected.
2026-10-06 15:12:48 +08:00
Tianyi Cui a3cb678875 Merge master current-directory support into subagent activations
Integrate the observed master checkpoint 7ab215e23e5d977aae7eb5d4fb91782d39969dbb without rewriting the earlier published checkpoints. Master adds current-directory events, directory-aware consumers, and retained Git worktrees; its old subagent startup paths overlapped the unified activation implementation.

Resolve directory selection within managed startup after acquiring parent ownership. Preserve the original Session origin while setting each new local child directory, retain persisted directory state on cold resume, and pass a required absolute cwd to external providers. Keep unified local/external ownership, full completion notices, parked-input host completion, SDK wait/error/shutdown behavior, and ACP cancellation semantics. Retain removal of the old in-process driver and continuation managers.

Combine SDK wait with the new directory requests, preserve the new master directory/worktree fixtures, regenerate derived catalogs from the combined sources, and refresh only the four stale prompt/schema sidecars. Preserve every protected historical schema, archived note, catalog-v2 acknowledgement, and recorded predecessor input.

Validation: full build; 59 CI static gates; 855 subagent tests with 100% coverage on changed core/external sources; affected consumer suites including 119 repaired-fixture tests; SDK 11 and headless 16 recorded scenarios; two local headless expected cases; six built nested SDK/ACP cases; lint, strict cast inventory, workspace constraints, and whitespace checks. Native Windows worker diagnostics remain additive and require the next CI run to classify the original crash.
2026-10-06 10:52:35 +08:00
Tianyi Cui 79bd3d8da7 feat(session): centralize working directories and directory context 2026-10-06 01:27:54 +08:00
Dudu-0223 3254ed611d Merge master and preserve catalog-recursive subagent discovery 2026-09-27 15:23:31 +08:00
07akioni 9d36f9c1fc refactor(telemetry): share reporting through a Cordis OTel service 2026-09-25 19:08:44 +08:00
07akioni 3dd52bfd9b fix(telemetry): own byte-bounded session request scheduling 2026-09-25 18:39:21 +08:00
07akioni 7b8c7dfff2 feat(telemetry): upload session log events through byte-bounded OTLP 2026-09-25 17:29:15 +08:00
lsdsjy 40da69ff75 refactor(account): isolate provider authentication and preserve model selections 2026-09-23 14:43:47 +08:00
creatixchu 4137d1d628 Merge remote-tracking branch 'origin/master' into fix/4775-multimodal-tool-retention
# Conflicts:
#	docs/config-catalog.i18n.yaml
#	docs/config-catalog.zh.md
#	docs/module-graph.i18n.yaml
#	docs/module-graph.md
#	docs/module-graph.zh.md
2026-09-22 10:52:21 +08:00
lsdsjy d5db0ff0c4 merge: sync master into desktop welcome branch 2026-09-21 19:53:28 +08:00
Turtle 601d6761e4 feat(settings): project volatile Config through profile-backed forms (#4587) 2026-09-21 18:26:03 +08:00
creatixchu ab102138c8 fix: retain ordered tool text and images within a token budget 2026-09-21 16:45:21 +08:00
Dudu-0223 307332baee Merge master and preserve compatible subagent catalog evolution 2026-09-20 17:35:18 +08:00
Tianyi Cui 99e22ebbeb refactor(llm): make the official DeepSeek adapter Messages-only 2026-09-20 14:37:19 +08:00
lsdsjy 68f8d32d83 fix(desktop,deepseek-account): share platform headers for Host API requests 2026-09-20 11:31:40 +08:00
lsdsjy 048297321a feat(desktop,credentials,client): integrate DeepSeek account sign-in and Platform pages 2026-09-20 11:31:40 +08:00
Dudu-0223 3d852a1cea refactor(subagent): unify local and external activation lifecycles 2026-09-17 14:54:10 +08:00
Dudu-0223 4565dafc58 refactor(subagent): default base fork delegation to continuable 2026-09-17 10:57:12 +08:00
Turtle abd765a600 refactor(hmr): own profile reload lifecycle through YAML 2026-09-15 13:01:06 +08:00
Turtle 438c862fea refactor: keep profile launch context data-only and HMR backend-scoped 2026-09-14 23:09:06 +08:00
Turtle d06e6b5519 feat: coordinate profile management through dsh-hmr 2026-09-14 23:09:06 +08:00
Turtle 98b92b683c feat: add current-profile plugin manager service and Web controls 2026-09-14 23:09:05 +08:00
creatixchu a7070008e7 Merge master and integrate image offload with both DeepSeek protocols 2026-09-14 10:19:11 +08:00
Tianyi Cui 2214891606 Merge pull request #4119 from deepseek-harness/worktree/workflow-sandbox-ptc
fix(workflow): execute orchestration through sandboxed PTC
2026-09-13 16:33:51 +08:00
Tianyi Cui 35af8698c2 fix(workflow): execute orchestration in the sandboxed PTC runtime 2026-09-13 14:47:56 +08:00
Tianyi Cui e08468954a feat(mcp): activate profile resource tools for configured servers 2026-09-13 11:52:42 +08:00
Tianyi Cui 73985344cd feat(presets): disable ralph in the default compositions 2026-09-13 00:15:39 +08:00
creatixchu 1faab27329 Merge origin/master into PR #3440 2026-09-10 13:48:38 +08:00
Yichen Jiang bc5fd3b8dc feat(llm): default Chat Completions to DeepSeek V41 Flash 2026-09-10 03:39:56 +08:00
creatixchu 4779f054b6 refactor(compaction): own image offload as a compaction executor
Move the IMAGE_OFFLOAD_REQUIRED recovery out of llm-retry into the new
@deepseek-ai/dsh-compaction-image-offload package, a sibling of the
tool-result pruner: image offload is compaction in another capacity
dimension. Each replaced node is priced through the seam's compaction/prune
shadow-price protocol before its marked copy lands. llm-retry returns to
master; the base bundle mounts the executor.
2026-09-08 15:10:16 +08:00
creatixchu fcb976d3dc refactor(llm): recover IMAGE_OFFLOAD_REQUIRED in llm-retry instead of a new plugin package
Remove the llm-image-offload package and every registration it needed. The
recovery branch lives in dsh-llm-retry, which already sits on the
agent/request-error waterfall: an IMAGE_OFFLOAD_REQUIRED failure advances the
image/offload watermark by the count the adapter names and retries the step
outside the provider retry budget. Drop pre-dispatch planning with it: routes
no longer declare a request-image budget on their model info, and the byte
checks that stay in the DeepSeek and pi-ai adapters are provisional until the
provider reports uncacheable images itself.
2026-09-07 17:03:23 +08:00
creatixchu 8c7b023d6c refactor(llm): move image offload planning into a plugin and trim the change surface
Keep the agent loop unchanged: the new @deepseek-ai/dsh-llm-image-offload
plugin plans the image/offload watermark on agent/pre-step from the routed
model's budget and advances it on agent/request-error when an adapter fails
with IMAGE_OFFLOAD_REQUIRED, returning the retry action the loop already
honours. Drop PreparedLlmCall.imageRequest and its validation, the planner
wrapper, the replay adapter's imageRequestMaxBytes option, and the duplicate
ACP snapshot scenario; share the exact-byte overflow count between the
DeepSeek and pi-ai adapters through requiredImageOffload(). The SDK image
snapshot replays an authored IMAGE_OFFLOAD_REQUIRED attempt instead of a
replay-only budget.
2026-09-07 15:58:20 +08:00
Tianyi Cui 40792330c0 fix(system-prompt): keep model persona prefix and place cwd in suffix 2026-09-06 20:35:42 +08:00
fz 17d8507b87 Merge origin/master into feat/sdk-disable-str-replace-editor 2026-09-05 23:57:47 +08:00
fz 36a4665144 feat(base): remove str_replace_editor from default tools 2026-09-05 23:37:13 +08:00
Tianyi Cui 9ffe85a512 OTel uploads explicit feedback for all users by default 2026-09-05 23:21:23 +08:00
ChinesezjcandTianyi Cui 02a029e679 Feedback is recorded in the session log and delivered through DeepSeek
Co-authored-by: Tianyi Cui <[email protected]>
2026-09-05 23:21:22 +08:00
fz 0a0f9e59ff feat(base): expose web fetch by default 2026-09-01 15:18:18 +08:00
fz ab9dcd5a2a Merge remote-tracking branch 'origin/master' into feat/sdk-default-web-fetch 2026-09-01 14:17:15 +08:00
fz cf7b0bd5a4 feat(headless): expose web fetch by default 2026-09-01 11:40:10 +08:00
Dudu-0223 b91e7ce366 Unify adjacent Agent messaging 2026-08-30 14:14:46 +08:00
Dudu-0223 7c626fb5d2 fix(subagent): gate model selection with explicit allowlist 2026-08-27 12:00:02 +08:00
creatixchu 9c37f7a553 Merge remote-tracking branch 'origin/master' into worktree/feedback-otel-enable
# Conflicts:
#	apps/cli/reference/README.i18n.yaml
#	apps/cli/reference/README.md
#	apps/cli/reference/README.zh.md
#	packages/bundle/base/cordis.patch.yml
2026-08-26 10:06:55 +08:00
creatixchu ac4a2f9792 fix(feedback): address review — accurate release wording, current-state notes, default-mode snapshot lane 2026-08-26 10:05:31 +08:00
_Kerman 3f34c026e3 Merge remote-tracking branch 'origin/master' into dshw/pr-deepseek-harness-deepseek-harness-2781 2026-08-25 20:04:32 +08:00