All pull requests

Pull requests

• fix(llmobs): skip thinking blocks in LLMJudge anthropic client [backport 4.16]

  #20893·dd-octo-sts[bot] opened Oct 8, 2026Bot·

  ·
  4

• feat(litellm): add usage and cost metrics for proxy calls

  #20892·elihschiff opened Oct 8, 2026Member·

  ·
  4

• ci(databases): add python 3.15 testing coverage

  Bits AIchangelog/no-changelog
  #20891·wconti27 opened Oct 8, 2026Contributor·

  ·
  7

• fix(tracing): keep W3C tracestate dd= member within 256 chars [DO NOT MERGE]

  #20887·mabdinur opened Oct 8, 2026Contributor·

  ·
  5

• fix(bootstrap): keep logging.captureWarnings working under ddtrace-run with gevent

  #20886·inchang-ing opened Oct 8, 2026·

  ·
  1

  2

• ci: test SSI Vault secrets with WIP templates

  changelog/no-changelog
  #20884·robertomonteromiguel opened Oct 8, 2026Contributor·

  ·
  5

• Fix #20783: include tool_use_prompt_token_count in google_genai LLM span tokens

  #20883·inchang-ing opened Oct 8, 2026·

  ·
  1

  1

• refactor(tracing): move sampling interfaces to native span data

  changelog/no-changelog
  #20882·florentinl opened Oct 8, 2026Contributor·

  ·
  14

  

• feat(mcp): support mcp 2.x

  #20881·timonschmidt opened Oct 8, 2026·

  ·
  2

  6

• fix(tracing): prevent quadratic query string obfuscation [backport 3.19]

  #20880·florentinl opened Oct 8, 2026Contributor·

  ·
  3

• Fix #20858: reset ASGI ExecutionContext between keep-alive requests

  #20877·inchang-ing opened Oct 8, 2026·

  ·
  1

  1

• Fix #20810: keep generated dd tracestate within the 256-character limit

  #20874·inchang-ing opened Oct 8, 2026·

  ·
  1

  1

• Fix #20782: map Anthropic 1h cache creation tokens to the correct span attribute

  #20873·inchang-ing opened Oct 8, 2026·

  ·
  1

  1

• fix(tracing): keep active APM_TRACING configs across remote config dispatches

  #20872·ebar0n opened Oct 8, 2026·

  ·
  1

  2

• fix(openai_agents): isolate concurrent workflow state

  #20870·yshapiro-57 opened Oct 8, 2026Contributor·

  ·
  3

• test: updated sdkrollout worker

  #20869·erikayasuda opened Oct 7, 2026Collaborator·

  ·
  5

• refactor(llmobs): abstract llmobs from mistralai integration code

  #20868·quinna-h opened Oct 7, 2026Contributor·

  ·
  4

• fix(langchain): join streamed chat chunks in linear time

  #20867·NarekA opened Oct 7, 2026·

  ·

• refactor(llmobs): abstract llmobs from llama_index integration code

  #20860·quinna-h opened Oct 7, 2026Contributor·

  ·
  4

• refactor(internal): don't release sys.monitoring tool if finalising

  changelog/no-changelogInternal
  #20856·KowalskiThomas opened Oct 7, 2026Collaborator·

  ·
  31

• chore(appsec): reuse shared distribution metadata

  changelog/no-changelog
  #20854·florentinl opened Oct 7, 2026Contributor·

  ·
  14

• feat(aiguard): evaluate openai-agents MCP server tool calls

  #20852·avara1986 opened Oct 7, 2026Member·

  ·
  5

• feat(aiguard): evaluate OpenAI hosted MCP tool calls

  #20850·avara1986 opened Oct 7, 2026Member·

  ·
  18

• fix(otel): use stable deployment environment resource key

  AI Generated
  #20847·mabdinur opened Oct 7, 2026Contributor·

  ·
  9

• fix(openai_agents): preserve interrupted workflow spans

  #20846·yshapiro-57 opened Oct 6, 2026Contributor·

  ·
  7