All pull requests

Pull requests

• chore(deps): bump the github-actions group across 1 directory with 3 updates

  dependenciesgithub_actions
  #2118·dependabot[bot] opened Nov 10, 2026Bot·

  ·

• fix(client): send additional_headers from async_api

  #2117·anubhavlal07 opened Nov 9, 2026·5 of 11 tasks·

  ·
  1

• feat(experiments): retries, timeouts, failure accounting and resume for experiments and batch evaluation

  #2115·Cookie0103 opened Nov 8, 2026·5 of 11 tasks·

  ·
  1

  16

• fix(prompts): skip cache write when cache_ttl_seconds is 0

  #2114·Mo0nl19ht opened Nov 8, 2026·6 of 11 tasks·

  ·
  1

• fix(batch-evaluation): separate stats for evaluators sharing a name

  #2113·Nafay-0 opened Nov 7, 2026·6 of 11 tasks·

  ·
  3

• fix(langchain): use the model name reported in the response

  #2112·Atishyy27 opened Nov 6, 2026·6 tasks done·

  ·
  1

  1

• fix(openai): record failed response stream status

  #2111·feiiiiii5 opened Nov 6, 2026·5 of 6 tasks·

  ·
  2

• fix(openai): end observations cancelled before a response

  #2110·feiiiiii5 opened Nov 6, 2026·5 of 6 tasks·

  ·
  4

• feat(api): update API spec from langfuse/langfuse 02962b3

  #2109·langfuse-bot opened Nov 6, 2026Collaborator·

  ·
  4

• fix(datasets): hydrate media under numeric object keys

  #2108·Aafimalek opened Nov 6, 2026·4 tasks done·

  ·

• fix(prompts): protect cache from stale background refreshes

  #2106·Aafimalek opened Nov 6, 2026·4 tasks done·

  ·
  2

• fix(observe): support async iterable streaming bodies

  #2105·Aafimalek opened Nov 6, 2026·4 tasks done·

  ·
  2

• fix(observe): resolve names for partial functions

  #2104·Aminark opened Nov 5, 2026·6 tasks done·

  ·
  1

• fix(evaluations): account for score duplication in evaluator stats and fix max_items deduplication

  #2103·viliamgonda opened Nov 5, 2026·

  ·
  1

  3

• fix(langchain): fallback to exception class name when error message is empty

  #2093·viliamgonda opened Nov 3, 2026·6 of 11 tasks·

  ·
  1

• docs(datasets): use result attributes in run evaluator example

  #2090·shuaiyafei opened Nov 3, 2026·4 tasks done·

  ·
  1

• fix(langchain): avoid double counting Anthropic prompt cache creation tokens

  #2088·viliamgonda opened Nov 1, 2026·6 of 11 tasks·

  ·
  1

  5

• fix(otel): widen the span queue when flush_at exceeds its default

  #2086·passionworkeer opened Nov 1, 2026·5 tasks done·

  ·
  1

  4

• fix(evaluations): flush scores without blocking the event loop

  #2085·6londe opened Nov 1, 2026·6 tasks done·

  ·
  1

• fix: preserve surrogate-containing metadata in OTLP

  #2084·DiadkoShmek opened Oct 31, 2026·

  ·
  4

• chore: make @hassiebp a code owner except for the Fern-generated API client

  #2081·hassiebp opened Oct 30, 2026Collaborator·4 tasks done·

  ·
  6

• fix(prompts): preserve tool messages in LangChain conversion

  #2080·feiiiiii5 opened Oct 30, 2026·6 of 11 tasks·

  ·
  7

• fix(prompts): honor empty fallback prompts

  #2079·feiiiiii5 opened Oct 30, 2026·6 of 11 tasks·

  ·
  2

• fix(openai): count base64 embedding vector dimensions

  #2077·as992949791 opened Oct 30, 2026·6 tasks done·

  ·
  4

• fix(prompts): accept fetched chat messages in create_prompt and fallback

  #2076·v0ropaev opened Oct 29, 2026·6 of 11 tasks·

  ·
  1