Overview

7 days30 days

Latest CI Pipeline Executions

Fuzzy

Succeeded
main
2b03b5c6 ci: Version Packages (#1281) Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
by github-act...
Succeeded
main
2b03b5c6 ci: Version Packages (#1281) Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
by github-act...
Succeeded
main
09bf8608 feat(ci): start AI review from label and approve waiting checks (#1280) * feat(ci): start AI review when the ai-review label is added * feat(ci): mark ready PRs secure and approve waiting workflows * feat(ci): print Grok review output in the Actions log * feat(ci): print each Grok review stream chunk in the Actions log * feat(ci): log review text, reasoning, and tool I/O * fix(ci): require a maintainer to trigger label review and fail closed on workflow changes
by Alem Tuzlak
Succeeded
main
09bf8608 feat(ci): start AI review from label and approve waiting checks (#1280) * feat(ci): start AI review when the ai-review label is added * feat(ci): mark ready PRs secure and approve waiting workflows * feat(ci): print Grok review output in the Actions log * feat(ci): print each Grok review stream chunk in the Actions log * feat(ci): log review text, reasoning, and tool I/O * fix(ci): require a maintainer to trigger label review and fail closed on workflow changes
by Alem Tuzlak
Succeeded
main
09bf8608 feat(ci): start AI review from label and approve waiting checks (#1280) * feat(ci): start AI review when the ai-review label is added * feat(ci): mark ready PRs secure and approve waiting workflows * feat(ci): print Grok review output in the Actions log * feat(ci): print each Grok review stream chunk in the Actions log * feat(ci): log review text, reasoning, and tool I/O * fix(ci): require a maintainer to trigger label review and fail closed on workflow changes
by Alem Tuzlak
Succeeded
main
4eb24ee4 chore: sync model metadata (#1264) Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
by github-act...
Succeeded
main
4eb24ee4 chore: sync model metadata (#1264) Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
by github-act...
Succeeded
main
4eb24ee4 chore: sync model metadata (#1264) Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
by github-act...
Succeeded
main
c3d4376b fix(ai-bedrock): forward usage from Converse structuredOutputStream() (#1278) * fix(ai-bedrock): forward usage from Converse structuredOutputStream() `BedrockConverseTextAdapter.structuredOutputStream()` iterated the Converse event stream handling only `contentBlockDelta` and `messageStop`. Converse reports token usage on a trailing `metadata` event, after the finish signal, so the terminal RUN_FINISHED was emitted with no `usage` and Bedrock consumers metering cost from it saw successful structured calls with no token signal. Capture usage from the `metadata` event and spread it onto RUN_FINISHED, mirroring `processConverseStream` in converse/stream-processor.ts and the non-stream `structuredOutput()` fixed in #1077. When no metadata event arrives the key stays absent, so unmetered stays distinguishable from zero-cost. Fixes #1276 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RgrmtdYU8TEo2kt59ZkJuY * test(ai-bedrock): cover usage on the Converse structured output stream Two cases in the existing StubAdapter suite, which drives the adapter's `sendStream` seam with canned Converse events. The first asserts that a trailing `metadata` event lands as `usage` on RUN_FINISHED. The second asserts that a stream without that event leaves the `usage` key absent, so an unmetered run stays distinguishable from a zero-cost one. The first case fails on c675499 without the fix in this PR. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RgrmtdYU8TEo2kt59ZkJuY --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
by Julius
Succeeded
main
c3d4376b fix(ai-bedrock): forward usage from Converse structuredOutputStream() (#1278) * fix(ai-bedrock): forward usage from Converse structuredOutputStream() `BedrockConverseTextAdapter.structuredOutputStream()` iterated the Converse event stream handling only `contentBlockDelta` and `messageStop`. Converse reports token usage on a trailing `metadata` event, after the finish signal, so the terminal RUN_FINISHED was emitted with no `usage` and Bedrock consumers metering cost from it saw successful structured calls with no token signal. Capture usage from the `metadata` event and spread it onto RUN_FINISHED, mirroring `processConverseStream` in converse/stream-processor.ts and the non-stream `structuredOutput()` fixed in #1077. When no metadata event arrives the key stays absent, so unmetered stays distinguishable from zero-cost. Fixes #1276 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RgrmtdYU8TEo2kt59ZkJuY * test(ai-bedrock): cover usage on the Converse structured output stream Two cases in the existing StubAdapter suite, which drives the adapter's `sendStream` seam with canned Converse events. The first asserts that a trailing `metadata` event lands as `usage` on RUN_FINISHED. The second asserts that a stream without that event leaves the `usage` key absent, so an unmetered run stays distinguishable from a zero-cost one. The first case fails on c675499 without the fix in this PR. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RgrmtdYU8TEo2kt59ZkJuY --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
by Julius
Succeeded
main
c3d4376b fix(ai-bedrock): forward usage from Converse structuredOutputStream() (#1278) * fix(ai-bedrock): forward usage from Converse structuredOutputStream() `BedrockConverseTextAdapter.structuredOutputStream()` iterated the Converse event stream handling only `contentBlockDelta` and `messageStop`. Converse reports token usage on a trailing `metadata` event, after the finish signal, so the terminal RUN_FINISHED was emitted with no `usage` and Bedrock consumers metering cost from it saw successful structured calls with no token signal. Capture usage from the `metadata` event and spread it onto RUN_FINISHED, mirroring `processConverseStream` in converse/stream-processor.ts and the non-stream `structuredOutput()` fixed in #1077. When no metadata event arrives the key stays absent, so unmetered stays distinguishable from zero-cost. Fixes #1276 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RgrmtdYU8TEo2kt59ZkJuY * test(ai-bedrock): cover usage on the Converse structured output stream Two cases in the existing StubAdapter suite, which drives the adapter's `sendStream` seam with canned Converse events. The first asserts that a trailing `metadata` event lands as `usage` on RUN_FINISHED. The second asserts that a stream without that event leaves the `usage` key absent, so an unmetered run stays distinguishable from a zero-cost one. The first case fails on c675499 without the fix in this PR. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RgrmtdYU8TEo2kt59ZkJuY --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
by Julius
Succeeded
main
845696b0 fix(gemini): preserve JSON Schema tool parameters (#1259) * fix(gemini): preserve JSON Schema tool parameters * test(e2e): repair Gemini tool-schema wire assertions The converter now sends tool schemas via `parametersJsonSchema`, which aimock's Gemini handler never records (it copies only `parameters` into its `/v1/_requests` journal). Two consequences, both fixed here. `provider-tool-dispatch-wire` asserted the journal's `function.parameters`, so it failed deterministically after the switch. Drop that matcher and the comment describing the old SDK-side upper-casing; the surviving `name: 'google_search'` check is what the spec relies on to prove the tool was not swallowed into `{googleSearch: {...}}`. The new Gemini case in `arktype-tool-wire` asserted only `ok: true`, which the route returns unless the SDK throws. Had the raw mount stopped matching, the request would fall through to aimock's native Gemini handler, return 200, and the test would pass while proving nothing. Assert on the mount's `Schema accepted` reply instead, which it emits only after verifying the request carried `parametersJsonSchema` with `unit.const` and no `parameters`. Verified: pointing the route at a model the mount does not intercept fails on the new assertion while `ok: true` still comes back. Claude-Session: https://claude.ai/code/session_01PG6uRwgdsxF9vrj973KsKX --------- Co-authored-by: Alem Tuzlak <t.zlak@hotmail.com> Co-authored-by: Tom Beckenham <34339192+tombeckenham@users.noreply.github.com>
by kolaworld
Succeeded
main
845696b0 fix(gemini): preserve JSON Schema tool parameters (#1259) * fix(gemini): preserve JSON Schema tool parameters * test(e2e): repair Gemini tool-schema wire assertions The converter now sends tool schemas via `parametersJsonSchema`, which aimock's Gemini handler never records (it copies only `parameters` into its `/v1/_requests` journal). Two consequences, both fixed here. `provider-tool-dispatch-wire` asserted the journal's `function.parameters`, so it failed deterministically after the switch. Drop that matcher and the comment describing the old SDK-side upper-casing; the surviving `name: 'google_search'` check is what the spec relies on to prove the tool was not swallowed into `{googleSearch: {...}}`. The new Gemini case in `arktype-tool-wire` asserted only `ok: true`, which the route returns unless the SDK throws. Had the raw mount stopped matching, the request would fall through to aimock's native Gemini handler, return 200, and the test would pass while proving nothing. Assert on the mount's `Schema accepted` reply instead, which it emits only after verifying the request carried `parametersJsonSchema` with `unit.const` and no `parameters`. Verified: pointing the route at a model the mount does not intercept fails on the new assertion while `ok: true` still comes back. Claude-Session: https://claude.ai/code/session_01PG6uRwgdsxF9vrj973KsKX --------- Co-authored-by: Alem Tuzlak <t.zlak@hotmail.com> Co-authored-by: Tom Beckenham <34339192+tombeckenham@users.noreply.github.com>
by kolaworld
Succeeded
main
9c9746bd feat(ai-gemini): add Gemini Omni 1.1 Flash (gemini-omni-1.1-flash) (#1273) * feat(ai-gemini): add Gemini Omni 1.1 Flash (gemini-omni-1.1-flash) Add the GA Interactions video model, keep gemini-omni-flash-preview as a deprecated alias until 2026-09-30, and map modelOptions.resolution onto response_format.resolution (360p | 720p | 1080p | 4k). Closes #1272 * fix(ai-gemini): put Omni resolution on size, not modelOptions Omni video size uses the same aspectRatio_resolution template as grok and byteplus (`16:9` or `16:9_1080p`). Drop the extra modelOptions.resolution field.
by Tom Beckenham
Succeeded
main
9c9746bd feat(ai-gemini): add Gemini Omni 1.1 Flash (gemini-omni-1.1-flash) (#1273) * feat(ai-gemini): add Gemini Omni 1.1 Flash (gemini-omni-1.1-flash) Add the GA Interactions video model, keep gemini-omni-flash-preview as a deprecated alias until 2026-09-30, and map modelOptions.resolution onto response_format.resolution (360p | 720p | 1080p | 4k). Closes #1272 * fix(ai-gemini): put Omni resolution on size, not modelOptions Omni video size uses the same aspectRatio_resolution template as grok and byteplus (`16:9` or `16:9_1080p`). Drop the extra modelOptions.resolution field.
by Tom Beckenham
Succeeded
main
9c9746bd feat(ai-gemini): add Gemini Omni 1.1 Flash (gemini-omni-1.1-flash) (#1273) * feat(ai-gemini): add Gemini Omni 1.1 Flash (gemini-omni-1.1-flash) Add the GA Interactions video model, keep gemini-omni-flash-preview as a deprecated alias until 2026-09-30, and map modelOptions.resolution onto response_format.resolution (360p | 720p | 1080p | 4k). Closes #1272 * fix(ai-gemini): put Omni resolution on size, not modelOptions Omni video size uses the same aspectRatio_resolution template as grok and byteplus (`16:9` or `16:9_1080p`). Drop the extra modelOptions.resolution field.
by Tom Beckenham
Succeeded
main
515b9b26 fix(ci): review PRs with grokBuildText grok-4.6 (#1279) * fix(ci): review PRs with grokBuildText instead of grokText * fix(ci): use grok-4.6 with grokBuildText for PR review * fix(ci): install Grok CLI in the review sandbox setup
by Alem Tuzlak
Succeeded
main
515b9b26 fix(ci): review PRs with grokBuildText grok-4.6 (#1279) * fix(ci): review PRs with grokBuildText instead of grokText * fix(ci): use grok-4.6 with grokBuildText for PR review * fix(ci): install Grok CLI in the review sandbox setup
by Alem Tuzlak
Succeeded
main
515b9b26 fix(ci): review PRs with grokBuildText grok-4.6 (#1279) * fix(ci): review PRs with grokBuildText instead of grokText * fix(ci): use grok-4.6 with grokBuildText for PR review * fix(ci): install Grok CLI in the review sandbox setup
by Alem Tuzlak
Succeeded
main
01f4ac7d ci: add in-process Grok PR review bot (#1277) * ci: add in-process Grok PR review bot Review open PRs with grok-4.6, comment, label, and optionally push listed polish. * ci: simplify Grok PR review bot Use chat() outputSchema for the verdict, import @tanstack/ai and @tanstack/ai-grok directly, and drop copied skills. * ci: harden Grok PR review bot against injection and path escape * ci: stop nested nx inside test:ai-review
by Alem Tuzlak