Issues

• [online evals] [evaluator] More intuitive project evaluator input mapping

  Arize-ai/phoenix#16901 · Filter by author cephalization opened on Oct 9, 2026
  

• [BUG]: experiments page times out on large datasets - compareExperiments SQL isn't paginated

  bug
  Arize-ai/phoenix#16891 · Filter by author elisevansbbfc opened on Oct 9, 2026
  1

• Add Mistral AI pricing support to Phoenix Cost Tracking

  c/tracesenhancement
  Arize-ai/phoenix#16878 · Filter by author salignatmoandal opened on Oct 8, 2026
  

• fix(agents): interrupting your own turn can briefly flip the chat into "busy elsewhere" mode

  Arize-ai/phoenix#16859 · Filter by author MoraVigoMalusardi opened on Oct 7, 2026
  1

• [tech spec] decision models within evaluators

  c/evals
  Arize-ai/phoenix#16852 · Filter by author cephalization opened on Oct 7, 2026
  

• [experiments] side-by-side tracing compare on experiment view

  Arize-ai/phoenix#16850 · Filter by author mikeldking opened on Oct 7, 2026
  

• [evals][python] Support the OpenAI Decisions API in classification evaluators

  c/evalsenhancementlanguage: python
  Arize-ai/phoenix#16846 · Filter by author mikeldking opened on Oct 7, 2026
  1

  

• [evals][ts] Support decision models (OpenAI Decisions API) in classification evaluators

  c/evalsenhancementlanguage: typescript
  Arize-ai/phoenix#16842 · Filter by author linear-code opened on Oct 7, 2026
  1

  

• docs: Weekly docs gap audit — 2026-10-07

  documentation
  Arize-ai/phoenix#16841 · Filter by author mikeldking opened on Oct 7, 2026

• Kubecon tutorial submission

  Arize-ai/phoenix#16840 · Filter by author linear-code opened on Oct 7, 2026

• Phoenix content blog

  Arize-ai/phoenix#16839 · Filter by author linear-code opened on Oct 7, 2026

• claude managed agents observability docs/cookbook

  Arize-ai/phoenix#16838 · Filter by author linear-code opened on Oct 7, 2026

• Persist column widths in the dataset examples and experiment compare tables

  good-agent-issue
  Arize-ai/phoenix#16833 · Filter by author juliendemarque-tulip opened on Oct 7, 2026
  1

• TS client: evaluator metadata is dropped when experiments record evaluations

  bugc/clientlanguage: typescriptpriority: high
  Arize-ai/phoenix#16832 · Filter by author juliendemarque-tulip opened on Oct 7, 2026
  2

  

• feat(mcp): route single-entity lookups to REST and aggregates to SQL in the MCP server instructions

  Arize-ai/phoenix#16820 · Filter by author axiomofjoy opened on Oct 7, 2026
  

• ci(harbor): add a workflow_dispatch runner for one Harbor job file

  Arize-ai/phoenix#16819 · Filter by author axiomofjoy opened on Oct 7, 2026
  

• feat(harbor): score which Phoenix API the agent used and default the judge to gpt-5.5

  Arize-ai/phoenix#16818 · Filter by author axiomofjoy opened on Oct 7, 2026
  

• Hill-climb Harbor verifiers measure the whole session and miss output-config annotation names

  Arize-ai/phoenix#16812 · Filter by author axiomofjoy opened on Oct 6, 2026
  

• [harbor] A job on a subset of tasks removes the other tasks' examples from the Phoenix dataset

  bugc/clientc/datasets
  Arize-ai/phoenix#16801 · Filter by author mikeldking opened on Oct 6, 2026
  

• [ENHANCEMENT]: store and show OTLP span links

  c/tracesenhancementlanguage: typescript
  Arize-ai/phoenix#16799 · Filter by author selimaytac opened on Oct 6, 2026
  1

  

• Record agent turn count and tool call count on Harbor runs

  Arize-ai/phoenix#16796 · Filter by author axiomofjoy opened on Oct 6, 2026
  

• [experiments] Metric charts on the experiment compare page

  c/experimentsc/metricsc/uienhancement
  Arize-ai/phoenix#16780 · Filter by author mikeldking opened on Oct 5, 2026
  

• feat(harbor): sql-benchmark and api-benchmark task sets with tool-usage diagnostics

  Arize-ai/phoenix#16777 · Filter by author axiomofjoy opened on Oct 5, 2026
  

• [ENHANCEMENT] Expose supported fields and formats when editing phoenix-client profiles

  Arize-ai/phoenix#16770 · Filter by author cephalization opened on Oct 5, 2026
  

• [BUG] [agents] UI SDK: Sequential navigation.goTo calls in a single script are flaky — second nav can silently fail

  bug
  Arize-ai/phoenix#16768 · Filter by author cephalization opened on Oct 4, 2026
  1