All pull requests

Pull requests

• test(dashboard): drop stale is_sis_compatibility_enabled mock

  #2949·YHC66 opened Oct 14, 2026Contributor·2 of 7 tasks·

  ·

• fix(ci): propagate OTEL failures from make test-unit

  #2947·Lothnic opened Oct 12, 2026·1 task done·

  ·
  6

• fix(core): preserve empty batch DataFrame columns

  #2946·LeonMAG opened Oct 12, 2026·3 of 5 tasks·

  ·
  6

• Document UNPARSABLE_SCORE and expose is_unparsable_score for integrators

  #2940·debu-sinha opened Oct 10, 2026Contributor·1 task done·

  ·
  1

  5

• fix(providers): record cost of Anthropic and Google feedback calls

  #2938·shehryars715 opened Oct 9, 2026·3 tasks done·

  ·
  1

  3

• docs(examples): compare a local Clef judge with Claude 5.5 judges

  #2936·sfc-gh-jreini opened Oct 8, 2026Contributor·4 tasks done·

  ·
  1

• fix(otel): trace async methods hidden behind a sync decorator

  #2934·Abelo9996 opened Oct 7, 2026Collaborator·3 of 7 tasks·

  ·
  1

• fix(otel): stop re-wrapping class methods for every new recorder

  #2928·Abelo9996 opened Oct 7, 2026Collaborator·3 of 7 tasks·

  ·
  1

  1

• test(otel): re-enable golden span tests that no longer vary by environment

  #2927·Abelo9996 opened Oct 7, 2026Collaborator·3 of 7 tasks·

  ·

• feat(feedback): add instruction retention conversation metric

  #2924·MONISMALIK1 opened Oct 6, 2026·1 task done·

  ·
  1

  13

• fix(anthropic): send sampling parameters via extra_body for anthropic>=1.0

  #2923·Zhuoxi2000 opened Oct 6, 2026Contributor·3 of 7 tasks·

  ·
  1

  2

• feat(feedback): add requirement satisfaction metric

  #2910·jstar0 opened Oct 4, 2026·1 task done·

  ·
  1

  13

• feat(feedback): add a test-tampering metric for coding-agent traces

  #2906·lorenzozanee opened Oct 3, 2026·3 of 7 tasks·

  ·
  1

  9

• feat(providers): add API Route feedback provider

  #2903·DennyHo0917 opened Oct 2, 2026·1 task done·

  ·
  5

• Add user_frustration conversation metric

  #2897·contentforge-press opened Oct 2, 2026·12 tasks done·

  ·
  10

• feat(feedback): emit gen_ai.evaluation.result events for metric results

  #2895·PranavD2905 opened Oct 1, 2026Contributor·4 of 7 tasks·

  ·
  1

  11

• Fix Metric argument precedence and with_arguments clobbering (#2865)

  #2874·HarshitR2004 opened Oct 1, 2026·3 of 7 tasks·

  ·
  1

  10

• fix(feedback): keep an entirely ungraded comprehensiveness run out of the average

  #2871·feiiiiii5 opened Oct 1, 2026Collaborator·3 of 7 tasks·

  ·
  8

• feat(feedback): add SelfConsistency intra-judge reliability wrapper

  #2868·shwetangsinha0509 opened Sep 30, 2026Contributor·2 tasks done·

  ·
  1

  22

• fix(feedback): sum child eval costs onto the EVAL_ROOT span

  #2863·Bdysj opened Sep 30, 2026Contributor·3 of 7 tasks·

  ·
  1

  5

• feat(apps): add TruAutoGen for AutoGen (AG2) multi-agent conversations

  #2802·desusaiteja opened Sep 19, 2026Contributor·4 of 7 tasks·

  ·
  16

• feat(cursor): recover token usage from the transcript on stop

  #2797·avanish-garg opened Sep 18, 2026Contributor·3 tasks done·

  ·
  1

  6

• feat(core): instrument official MCP client calls

  #2791·adity982 opened Sep 15, 2026·1 task done·

  ·
  1

  7

• fix(hotspots): point README at the actual Streamlit module path

  #2777·dajiaohuang opened Sep 11, 2026·1 task done·

  ·
  1

• Add trulens-connectors-openeval: to_openeval()/from_openeval() for EvalPort interchange

  #2757·adhabnr-ux opened Sep 2, 2026·

  ·
  8