Overview - Ai

aiWorkspace

CI Pipeline Executions
- Filtered
Runs
Demo

Insights

Compare tasks
Analytics

Overview

ai

Loading workspace stats

Loading workspace insights...

Statistics interval

7 days30 days

Latest CI Pipeline Executions

Filter

Fuzzy

Sort by

© 2026 - Nx Cloud

Terms of Service Privacy Policy Changelog Status Docs Contact Nx Cloud Pricing Company @NxDevTools

aiWorkspace

Average execution time

9m 48s

949 CI pipeline executions

Computation time saved

22d 2h

Cache hit rate

56%

Time to Green

35m 38s

Was 39m 23s, improved by 10%

Succeeded
feat/boxd-sandbox-provider
1ea24da1 fix(ai): don't duplicate the reasoning block when a run is rejoined mid-stream (#1366) * fix(ai): don't duplicate the reasoning block when a run is rejoined mid-stream Fixes #1344 * fix(ai): only adopt a thinking part that has no stepId field --------- Co-authored-by: Alem Tuzlak <t.zlak@hotmail.com>
by citizen204
Succeeded
feat/ai-sandbox-e2b
e58f6bf2 fix(ai): don't duplicate the reasoning block when a run is rejoined mid-stream (#1366) * fix(ai): don't duplicate the reasoning block when a run is rejoined mid-stream Fixes #1344 * fix(ai): only adopt a thinking part that has no stepId field --------- Co-authored-by: Alem Tuzlak <t.zlak@hotmail.com>
by citizen204
Succeeded
main
abb0169b fix(ai): don't duplicate the reasoning block when a run is rejoined mid-stream (#1366) * fix(ai): don't duplicate the reasoning block when a run is rejoined mid-stream Fixes #1344 * fix(ai): only adopt a thinking part that has no stepId field --------- Co-authored-by: Alem Tuzlak <t.zlak@hotmail.com>
by citizen204
Succeeded
main
abb0169b fix(ai): don't duplicate the reasoning block when a run is rejoined mid-stream (#1366) * fix(ai): don't duplicate the reasoning block when a run is rejoined mid-stream Fixes #1344 * fix(ai): only adopt a thinking part that has no stepId field --------- Co-authored-by: Alem Tuzlak <t.zlak@hotmail.com>
by citizen204
Succeeded
fix-structured-output-length-finish
c4700afc fix(ai): don't duplicate the reasoning block when a run is rejoined mid-stream (#1366) * fix(ai): don't duplicate the reasoning block when a run is rejoined mid-stream Fixes #1344 * fix(ai): only adopt a thinking part that has no stepId field --------- Co-authored-by: Alem Tuzlak <t.zlak@hotmail.com>
by citizen204
Succeeded
fix/compatible-summary-token-cap
0d923653 fix(ai-client): surface persistence:true mount hydration failures (#1354) * fix(ai-client): surface persistence:true mount hydration failures A failed server-driven thread load (connection.hydrate) was swallowed by a bare `catch { return }` in `hydrateFromServer`: onError never fired, error stayed undefined, and status stayed 'ready' with zero messages, so an app could not tell a broken load from an empty thread. Add a `failHydration` path mirroring GenerationClient (status 'error' + error + onError), and route ByokMissingError / locked ByokBlockedError to the key-request flow like the send path does. Genuine misses and post-detach failures stay silent. Fixes #1331 * test(ai-client): clear the hydration error after a successful retry * test(ai-client): type the hydration error tests against main's adapter ResumableConnectConnectionAdapter now requires joinRun. --------- Co-authored-by: Alem Tuzlak <t.zlak@hotmail.com>
by citizen204
Succeeded
fix/anthropic-interleaved-thinking-order
caffc33e fix(ai-client): surface persistence:true mount hydration failures (#1354) * fix(ai-client): surface persistence:true mount hydration failures A failed server-driven thread load (connection.hydrate) was swallowed by a bare `catch { return }` in `hydrateFromServer`: onError never fired, error stayed undefined, and status stayed 'ready' with zero messages, so an app could not tell a broken load from an empty thread. Add a `failHydration` path mirroring GenerationClient (status 'error' + error + onError), and route ByokMissingError / locked ByokBlockedError to the key-request flow like the send path does. Genuine misses and post-detach failures stay silent. Fixes #1331 * test(ai-client): clear the hydration error after a successful retry * test(ai-client): type the hydration error tests against main's adapter ResumableConnectConnectionAdapter now requires joinRun. --------- Co-authored-by: Alem Tuzlak <t.zlak@hotmail.com>
by citizen204
Succeeded
main
a620c90d fix(ai-client): surface persistence:true mount hydration failures (#1354) * fix(ai-client): surface persistence:true mount hydration failures A failed server-driven thread load (connection.hydrate) was swallowed by a bare `catch { return }` in `hydrateFromServer`: onError never fired, error stayed undefined, and status stayed 'ready' with zero messages, so an app could not tell a broken load from an empty thread. Add a `failHydration` path mirroring GenerationClient (status 'error' + error + onError), and route ByokMissingError / locked ByokBlockedError to the key-request flow like the send path does. Genuine misses and post-detach failures stay silent. Fixes #1331 * test(ai-client): clear the hydration error after a successful retry * test(ai-client): type the hydration error tests against main's adapter ResumableConnectConnectionAdapter now requires joinRun. --------- Co-authored-by: Alem Tuzlak <t.zlak@hotmail.com>
by citizen204
Succeeded
main
a620c90d fix(ai-client): surface persistence:true mount hydration failures (#1354) * fix(ai-client): surface persistence:true mount hydration failures A failed server-driven thread load (connection.hydrate) was swallowed by a bare `catch { return }` in `hydrateFromServer`: onError never fired, error stayed undefined, and status stayed 'ready' with zero messages, so an app could not tell a broken load from an empty thread. Add a `failHydration` path mirroring GenerationClient (status 'error' + error + onError), and route ByokMissingError / locked ByokBlockedError to the key-request flow like the send path does. Genuine misses and post-detach failures stay silent. Fixes #1331 * test(ai-client): clear the hydration error after a successful retry * test(ai-client): type the hydration error tests against main's adapter ResumableConnectConnectionAdapter now requires joinRun. --------- Co-authored-by: Alem Tuzlak <t.zlak@hotmail.com>
by citizen204
Succeeded
fix-1331-hydration-error
c369b354 Merge c79aa58f2901e1f62594e3c7b6f8ead4105f9f24 into a0f7c14a9d9a4b2e72e87b976f46d193deb5921b
by citizen204
Succeeded
fix-1344-resume-thinking-dup
7a986533 Merge f38fbd03c5f17f9b01f37a3028a708754ea8a1bc into a0f7c14a9d9a4b2e72e87b976f46d193deb5921b
by citizen204
Succeeded
main
418142ba fix(ai-gemini): dedupe functionResponse parts by id, not name (#1198) * fix(ai-gemini): dedupe functionResponse parts by id, not name Two parallel calls to the same tool share a functionResponse.name but have distinct ids (msg.toolCallId, already set at both construction sites). mergeConsecutiveSameRoleMessages deduped by name, so the second response to a repeated same-tool call was dropped, leaving Gemini with fewer response parts than call parts on the next request: 400 INVALID_ARGUMENT: Please ensure that the number of function response parts is equal to the number of function call parts of the function call turn. Key the dedup on functionResponse.id instead — it still collapses a genuine duplicate tool result (same id twice), and now also preserves both responses when the model fires the same tool twice in one turn. Fixes #894 * chore: add changeset for #894 fix * ci: apply automated fixes --------- Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
by citizen204
Succeeded
main
418142ba fix(ai-gemini): dedupe functionResponse parts by id, not name (#1198) * fix(ai-gemini): dedupe functionResponse parts by id, not name Two parallel calls to the same tool share a functionResponse.name but have distinct ids (msg.toolCallId, already set at both construction sites). mergeConsecutiveSameRoleMessages deduped by name, so the second response to a repeated same-tool call was dropped, leaving Gemini with fewer response parts than call parts on the next request: 400 INVALID_ARGUMENT: Please ensure that the number of function response parts is equal to the number of function call parts of the function call turn. Key the dedup on functionResponse.id instead — it still collapses a genuine duplicate tool result (same id twice), and now also preserves both responses when the model fires the same tool twice in one turn. Fixes #894 * chore: add changeset for #894 fix * ci: apply automated fixes --------- Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
by citizen204
Succeeded
main
7c4b25eb fix(ai): ignore a repeat TOOL_CALL_START for an already-tracked toolCallId (#1197) * fix(ai): ignore a repeat TOOL_CALL_START for an already-tracked toolCallId AG-UI's TOOL_CALL_START carries no index, and a non-first-party or malformed producer can send a second START for a toolCallId that is already being tracked by ToolCallManager. Without a guard: - a repeat with the same index overwrites the slot, wiping any TOOL_CALL_ARGS already accumulated (arguments reset to '') - a repeat with a missing/different index inserts a duplicate map row, so getToolCalls() returns the id twice and the engine runs the tool twice Skip the repeat instead. First-party adapters emit START only once per call, so this only changes behavior for custom/malformed streams. Fixes #1187 * chore: add changeset for #1187 fix * ci: apply automated fixes --------- Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
by citizen204
Succeeded
main
7c4b25eb fix(ai): ignore a repeat TOOL_CALL_START for an already-tracked toolCallId (#1197) * fix(ai): ignore a repeat TOOL_CALL_START for an already-tracked toolCallId AG-UI's TOOL_CALL_START carries no index, and a non-first-party or malformed producer can send a second START for a toolCallId that is already being tracked by ToolCallManager. Without a guard: - a repeat with the same index overwrites the slot, wiping any TOOL_CALL_ARGS already accumulated (arguments reset to '') - a repeat with a missing/different index inserts a duplicate map row, so getToolCalls() returns the id twice and the engine runs the tool twice Skip the repeat instead. First-party adapters emit START only once per call, so this only changes behavior for custom/malformed streams. Fixes #1187 * chore: add changeset for #1187 fix * ci: apply automated fixes --------- Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
by citizen204
Succeeded
main
a54ed1f4 fix(ai-openai): stream reasoning from OpenAI-compatible providers (#1367) Fixes #982
by citizen204
Succeeded
main
a54ed1f4 fix(ai-openai): stream reasoning from OpenAI-compatible providers (#1367) Fixes #982
by citizen204
Succeeded
fix-1187-tool-call-manager-repeat-start
1f3a6b86 Merge 53394ccda27f453935ed1823f7be13c4b5be2c26 into bad612a3ae64702a3194c986a70f1905cad42751
by citizen204
Succeeded
fix-894-gemini-parallel-tool-dedup
2b48ba86 Merge 7ec18ee24c60ddf6b5f73ccc7d4703fd5e20a99d into bad612a3ae64702a3194c986a70f1905cad42751
by citizen204
Succeeded
fix-982-compatible-reasoning
bbba82b5 fix(ai-openai): stream reasoning from OpenAI-compatible providers Fixes #982
by citizen204