[bot] Instrument Z.AI Java SDK (34,271 weekly downloads)
Dieses Issue hat noch niemand übernommen.
Bewertung
- Schwierigkeit
- 4/5
- Geschätzter Aufwand
- 3-5 Tage
- Anfängerfreundlichkeit
- 45/100
- Issue-Typ
- Feature
- Klarheit
- Größtenteils klar
- Aktivitätsstatus
- Aktiv
- Tech-Stack
- java
- Bereich
- backend, observability
Rechercherichtung
Read AGENTS.md and compare the existing modules under braintrust-sdk/instrumentation/*, especially their build.gradle files. Then inspect the upstream zai-org/z-ai-sdk-java client APIs and the repository’s instrumentation patterns. Done means the Java SDK’s relevant Z.AI calls are traced and covered by tests; the issue does not identify specific entry points or tests.
Vom Indexierungsmodell aus dem Issue-Text verfasst.
Beschreibung
Summary
Z.AI (formerly Zhipu AI, makers of the GLM model family) publishes an official Java SDK, zai-org/z-ai-sdk-java, distributed on Maven Central as ai.z.openapi:zai-sdk. It exposes a dedicated client (ZaiClient / ZhipuAiClient) for execution-oriented generative AI calls — chat completions (sync + streaming), function/tool calling, embeddings, vision, image generation, video generation, and speech. This SDK is not covered by any existing instrumentation module in this repo, and its client classes are entirely distinct from the com.openai/com.anthropic/dev.langchain4j/Spring AI/Bedrock/Google GenAI classes already instrumented, so none of today's bytecode instrumentation applies to it.
What is missing
| Capability | Upstream (ai.z.openapi:zai-sdk) |
Braintrust Java SDK instrumentation |
|---|---|---|
| Chat completions (sync) | ✅ ChatCompletionService |
❌ not instrumented |
| Streaming chat completions | ✅ Flowable-based streaming |
❌ not instrumented |
| Function / tool calling | ✅ | ❌ not instrumented |
| Embeddings | ✅ | ❌ not instrumented |
| Vision / multimodal | ✅ | ❌ not instrumented |
| Image / video generation | ✅ | ❌ not instrumented |
None of braintrust-sdk/instrumentation/* targets the ai.z.openapi group — the directory currently contains only anthropic_2_2_0, aws_bedrock_2_30_0, genai_1_18_0, langchain_1_14_0, langchain_1_8_0, openai_2_15_0, springai_1_0_0, and springai_2_0_0.
Weekly downloads
Maven Central does not publish public per-artifact download counts, so per this repo's established precedent (e.g. #162) the primary tracked distribution for this SDK family is cited instead: the official Python package zai-sdk on PyPI had 34,271 downloads in the last 7 days, as of 2026-10-19, per pypistats.org (https://pypistats.org/api/packages/zai-sdk/recent → {"last_week": 34271}; corroborated by the daily overall series at https://pypistats.org/api/packages/zai-sdk/overall covering 2026-04-08–2026-10-18).
Braintrust docs status
Checked braintrust.dev/docs/admin/ai-providers: Braintrust lists OpenAI, Anthropic, Google Vertex AI, Azure AI Foundry, Bedrock, and custom providers as configurable AI providers. GLM (Z.AI's model family) appears only as a Braintrust-hosted built-in model (GLM-5.2, GLM-5.3 Flash, served via Baseten through the Braintrust gateway using an OpenAI-compatible endpoint) — there is no direct Z.AI/Zhipu provider integration or SDK-instrumentation guidance. A user calling Z.AI's own API directly via the official zai-sdk Java client (rather than routing through Braintrust's gateway with the OpenAI SDK) gets no tracing today.
Upstream sources
- SDK repo: https://github.com/zai-org/z-ai-sdk-java
- Official docs: https://docs.z.ai/guides/develop/java/introduction
- Maven Central: https://central.sonatype.com/artifact/ai.z.openapi/zai-sdk (latest
0.3.5; active release cadence from0.0.6through0.3.5across 2025–2026)
Braintrust docs sources
- https://www.braintrust.dev/docs/admin/ai-providers
- https://braintrust.dev/blog/glm-52-vs-opus-48-technical-report
Local files inspected
AGENTS.mdbraintrust-sdk/instrumentation/*/build.gradle(anthropic_2_2_0, aws_bedrock_2_30_0, genai_1_18_0, langchain_1_14_0, langchain_1_8_0, openai_2_15_0, springai_1_0_0, springai_2_0_0)- Directory listing of
braintrust-sdk/instrumentation/*
- Vorherrschende Sprache
- Java
- Sterne
- 21
- Forks
- 5
- Ø Merge
- 18 Std. 29 Min.
- Gemergte PRs (30 T.)
- 8
Entwicklungsumgebung
- Kein Dockerfile und keine Docker-Compose-Datei
- Keine Pull-Request-Vorlage
- Beitragsleitfaden lesen
Erste Schritte
- Lesen Sie das ganze Issue und danach den Beitragsleitfaden des Projekts.
- Schreiben Sie ins Issue, dass Sie es übernehmen — das erspart doppelte Arbeit.
- Forken Sie das Repository und arbeiten Sie in einem Branch.
- Öffnen Sie einen Pull Request, der die Issue-Nummer nennt.
Mehr aus braintrustdata/braintrust-sdk-java
-
Schwierigkeit 2/5 1-3 Stunden Anfängerfreundlichkeit 82/100
-
Schwierigkeit 2/5 1-3 Stunden Anfängerfreundlichkeit 68/100
-
Schwierigkeit 5/5 Über eine Woche Anfängerfreundlichkeit 35/100
-
Schwierigkeit 4/5 3-5 Tage Anfängerfreundlichkeit 55/100
-
Schwierigkeit 4/5 3-5 Tage Anfängerfreundlichkeit 38/100
Alle Issues in braintrustdata/braintrust-sdk-java
Ähnliche Issues
-
Schwierigkeit 2/5 1-3 Stunden Anfängerfreundlichkeit 79/100
checkstyle/checkstyle#22145 ·
Maintainer antworten meist innerhalb von 1 Tag
-
rset command throws NPE in ResultsFormat.guessSyntaxEvtl. vergeben @Ostrzyciel hat das heute übernommen. Offenbug
Schwierigkeit 2/5 1-3 Stunden Anfängerfreundlichkeit 72/100
Maintainer antworten meist innerhalb von 1 Tag
-
enhancement good first issue help wanted
Schwierigkeit 2/5 1-3 Stunden Anfängerfreundlichkeit 83/100
Maintainer antworten meist innerhalb von 1 Tag
-
Schwierigkeit 2/5 1-3 Stunden Anfängerfreundlichkeit 82/100
objectionary/eo#9593 ·
Maintainer antworten meist innerhalb von 1 Tag
-
enhancement
Schwierigkeit 2/5 1-3 Stunden Anfängerfreundlichkeit 65/100