v6.14.0 proposal

[dd-octo-sts]

Features

• apm: Add data pipeline-backed agentless mode #9839
• AppSec: Extract API Security schemas in Lambda #10014
• Crash Tracking: Support agentless intake #10121
• Dynamic Instrumentation: Support agentless intake #9881
• General: Add async Orchestrion context callbacks #10080
• General: Await callbacks at function start #10123
• General: Enable client self-telemetry #9698
• graphql: Trace graphql-jit executions #9343
• knex, sequelize: Add pool acquire span #8924
• Serverless: Add AWS MicroVM support #9709
• Test Optimization: Correlate WebdriverIO tests with RUM #10049
• Test Optimization: Upload browser test failure videos #9927
• Test Optimization: Upload WebdriverIO failure screenshots #10113

Fixes

• AppSec: Normalize sampling priority in the API Security sampler #9953
• esbuild: Resolve imports in built-in subpath wrappers #10102
• fs: Preserve stream constructor errors #10129
• General: Pass raw container ID to trace intake #10105
• General: Prefer stable OTel deployment environment #10050
• General: Preserve explicit telemetry configuration #10124
• General: Preserve OTel metrics in agentless mode #10142
• General: Preserve RegExp sampling rule matchers #10134
• General: Return empty RUM data without an active span #10130
• graphql: Record collapsed JIT resolver errors #10027
• graphql: Record reused non-JIT resolver errors #10082
• LLM Observability: Do not mark graph interrupt errors as actual errors #10010
• LLM Observability: Prevent duplicate tags in x-datadog-tags on injection #10100
• LLM Observability: Properly patch api promise prototypes for complete coverage #9951
• LLM Observability: Resolve unbounded memory growth for the vercel ai sdk llmobs plugin #10106
• mongodb: Extract queries from document sequences #10046
• net: Clean up synchronous connection failures #10128
• next: Trace compiled response lifecycles #9708
• Serverless: Coalesce overlapping Vercel telemetry flushes #9954
• Test Optimization: Bound payload delivery lifecycle #10044
• Test Optimization: Cache settings request in the filesystem cache #9989
• Test Optimization: Cap screenshot upload retries #10056
• Test Optimization: Extend final flush timeout #10040
• Test Optimization: Increase payload request concurrency #10038
• Test Optimization: Restore pre-defer suite reporting #10017
• Test Optimization: Retry transient HTTP responses #10039
• Test Optimization: Support Mocha 12 #10095
• Test Optimization: Support Vitest 5 #10161
• Test Optimization: Use a 10-second flush interval #10018
• undici: Prevent context retention after requests #9406

Performance

• graphql: Reduce inactive resolver overhead #10136
• LLM Observability: Build provider strings directly #10088
• rewriter: Defer transformer load to first rewrite #10099
• Test Optimization: Build formatted strings directly #10087
• trace-encoder: Cache stable strings across payloads #10057
• tracer: Avoid conditional object spread allocations #10052
• tracer: Build runtime strings directly #10089

Documentation

• agents: Simplify repository guidance #9969

Internal (CI, Testing, Benchmarking)

• bun: Expand runtime smoke coverage #10048
• Dependencies: Bump protobufjs from 8.7.2 to 8.8.0 in /vendor in the vendor-minor-and-patch-dependencies group across 1 directory #10107
• dns: Remove live resolver dependency #9950
• Dynamic Instrumentation: Replace axios with fetch #10053
• eslint: Detect direct literal joins #9926
• eslint: Detect guarded array joins #10140
• eslint: Enforce direct string construction #10090
• eslint: Name returned functions before invocation #10083
• eslint: Prohibit immediate call-result invocation #10060
• General: Fix v5 release proposal failures #10143
• General: Reduce system tests execution time, split AppSec ubuntu matrix, rework All Green report uploads #9197
• install: Pin standardwebhooks for Anthropic fixtures #10054
• instrumentations: Use supported Confluent Node versions #9771
• lint: Enable eslint-plugin-regexp rules #10141
• mariadb: Limit clock stubs to synchronous calls #10077
• sirun: Make benchmark workloads deterministic #9715
• Test Optimization: Build config wrappers directly #10059
• Test Optimization: Cover missing source test reports #9931
• Test Optimization: Cover webdriverio browser provider #10116
• Test Optimization: Support Cypress 16 #10114

Contributors

[dd-octo-sts]

Overall package size

Self size: 8.83 MB
Deduped: 9.49 MB
No deduping: 9.49 MB

Dependency sizes

| name | version | self size | total size | |------|---------|-----------|------------| | import-in-the-middle | 3.4.0 | 127.33 kB | 447.04 kB | | opentracing | 0.14.7 | 194.81 kB | 194.81 kB | | dc-polyfill | 0.1.11 | 25.74 kB | 25.74 kB |

_{🤖 This report was automatically generated by heaviest-objects-in-the-universe}

[datadog-official]

🎉 All green!

🧪 All tests passed
❄️ No new flaky tests detected

🎯 Code Coverage (details)
• Patch Coverage: 99.15%
• Overall Coverage: 98.02%

_{This comment will be updated automatically if new data arrives.
🔗 Commit SHA: 4d60d8d | Docs | View more details | Give us feedback!}

[codecov]

Codecov Report

❌ Patch coverage is 99.55011% with 18 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (v6.x@298497a). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
...instrumentations/src/mocha/webdriverio-protocol.js	88.77%	11 Missing ⚠️
...ckages/datadog-instrumentations/src/vitest-main.js	95.65%	4 Missing ⚠️
...ckages/datadog-instrumentations/src/webdriverio.js	99.68%	3 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             v6.x   #10068   +/-   ##
=======================================
  Coverage        ?   98.64%           
=======================================
  Files           ?     1004           
  Lines           ?   157369           
  Branches        ?       74           
=======================================
  Hits            ?   155233           
  Misses          ?     2136           
  Partials        ?        0           

Flag	Coverage Δ
ai-guard	65.05% <58.98%> (?)
apm-capabilities	63.05% <63.37%> (?)
apm-integrations	81.41% <95.75%> (?)
appsec	58.70% <54.51%> (?)
debugger	69.84% <91.30%> (?)
instrumentation	53.72% <59.56%> (?)
llmobs	78.98% <61.01%> (?)
openfeature	66.55% <87.29%> (?)
platform	69.30% <91.30%> (?)
profiling	66.77% <87.29%> (?)
serverless	65.73% <60.56%> (?)
test-optimization	82.72% <85.96%> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[pr-commenter]

Benchmarks

Benchmark execution time: 2026-09-04 15:57:50

Comparing candidate commit 4d60d8d in PR branch v6.14.0-proposal with baseline commit 298497a in branch v6.x.

📊 Benchmarking dashboard

Found 20 performance improvements and 1 performance regressions! Performance is the same for 2271 metrics, 21 unstable metrics.

Explanation

This is an A/B test comparing a candidate commit's performance against that of a baseline commit. Performance changes are noted in the tables below as:

• 🟩 = significantly better candidate vs. baseline
• 🟥 = significantly worse candidate vs. baseline

We compute a confidence interval (CI) over the relative difference of means between metrics from the candidate and baseline commits, considering the baseline as the reference.

If the CI is entirely outside the configured SIGNIFICANT_IMPACT_THRESHOLD (or the deprecated UNCONFIDENCE_THRESHOLD), the change is considered significant.

Feel free to reach out to #apm-benchmarking-platform on Slack if you have any questions.

More details about the CI and significant changes

You can imagine this CI as a range of values that is likely to contain the true difference of means between the candidate and baseline commits.

CIs of the difference of means are often centered around 0%, because often changes are not that big:

---------------------------------(------|---^--------)-------------------------------->
                              -0.6%    0%  0.3%     +1.2%
                                 |          |        |
         lower bound of the CI --'          |        |
sample mean (center of the CI) -------------'        |
         upper bound of the CI ----------------------'

As described above, a change is considered significant if the CI is entirely outside the configured SIGNIFICANT_IMPACT_THRESHOLD (or the deprecated UNCONFIDENCE_THRESHOLD).

For instance, for an execution time metric, this confidence interval indicates a significantly worse performance:

----------------------------------------|---------|---(---------^---------)---------->
                                       0%        1%  1.3%      2.2%      3.1%
                                                  |   |         |         |
       significant impact threshold --------------'   |         |         |
                      lower bound of CI --------------'         |         |
       sample mean (center of the CI) --------------------------'         |
                      upper bound of CI ----------------------------------'

scenario:encoders-0.4-20

• 🟩 max_rss_usage [-14.340MB; -4.711MB] or [-10.305%; -3.386%]

scenario:encoders-0.4-events-native-20

• 🟥 max_rss_usage [+3.199MB; +9.671MB] or [+2.268%; +6.857%]

scenario:encoders-0.4-immediate-flush-20

• 🟩 cpu_user_time [-1.815s; -0.510s] or [-10.530%; -2.961%]
• 🟩 execution_time [-1.869s; -0.515s] or [-10.787%; -2.970%]
• 🟩 instructions [-13.8G instructions; -4.1G instructions] or [-9.178%; -2.707%]
• 🟩 throughput [+512.341op/s; +1881.285op/s] or [+2.943%; +10.806%]

scenario:encoders-0.4-immediate-flush-24

• 🟩 cpu_user_time [-1507.772ms; -417.031ms] or [-12.186%; -3.370%]
• 🟩 execution_time [-1506.290ms; -409.365ms] or [-12.105%; -3.290%]
• 🟩 instructions [-13.0G instructions; -3.8G instructions] or [-11.115%; -3.230%]
• 🟩 throughput [+788.255op/s; +2946.860op/s] or [+3.247%; +12.138%]

scenario:encoders-0.4-immediate-flush-26

• 🟩 cpu_user_time [-1065.252ms; -301.708ms] or [-9.020%; -2.555%]
• 🟩 execution_time [-1053.253ms; -275.865ms] or [-8.835%; -2.314%]
• 🟩 instructions [-9.8G instructions; -2.9G instructions] or [-8.247%; -2.427%]
• 🟩 throughput [+572.582op/s; +2246.827op/s] or [+2.266%; +8.892%]

scenario:plugin-graphql-long-with-depth-off-26

• 🟩 cpu_user_time [-487.874ms; -149.890ms] or [-6.580%; -2.022%]
• 🟩 execution_time [-520.952ms; -160.712ms] or [-6.895%; -2.127%]
• 🟩 throughput [+7.409op/s; +23.897op/s] or [+2.179%; +7.029%]

scenario:plugin-graphql-long-with-depth-on-max-20

• 🟩 instructions [-4.0G instructions; -1.5G instructions] or [-11.240%; -4.254%]

scenario:plugin-graphql-long-with-depth-on-max-26

• 🟩 cpu_user_time [-746.931ms; -243.801ms] or [-9.714%; -3.171%]
• 🟩 execution_time [-782.972ms; -258.373ms] or [-10.003%; -3.301%]
• 🟩 throughput [+7.638op/s; +23.620op/s] or [+3.259%; +10.078%]

Unstable benchmarks

These benchmarks have a confidence interval too wide to call a change; treat them as noise rather than signal.

scenario:appsec-iast-with-vulnerability-iast-enabled-default-config-20

• unstable max_rss_usage [-24.946MB; +14.786MB] or [-6.816%; +4.040%]

scenario:debugger-line-probe-with-snapshot-minimal-20

• unstable max_rss_usage [-5.986MB; +13.108MB] or [-3.745%; +8.201%]

scenario:encoders-0.4-immediate-flush-20

• unstable max_rss_usage [-13.741MB; +2.914MB] or [-17.366%; +3.683%]

scenario:plugin-cassandra-driver-batch-20

• unstable cpu_user_time [-439.140ms; -185.946ms] or [-18.550%; -7.854%]
• unstable execution_time [-439.479ms; -186.019ms] or [-18.522%; -7.840%]
• unstable instructions [-4.6G instructions; -1.8G instructions] or [-17.553%; -7.036%]
• unstable throughput [+64730.547op/s; +157297.095op/s] or [+7.481%; +18.179%]

scenario:plugin-cassandra-driver-batch-24

• unstable cpu_user_time [-483.178ms; -200.198ms] or [-19.177%; -7.946%]
• unstable execution_time [-488.861ms; -196.752ms] or [-19.228%; -7.739%]
• unstable instructions [-4.5G instructions; -1.9G instructions] or [-17.428%; -7.316%]
• unstable throughput [+61836.389op/s; +155710.798op/s] or [+7.686%; +19.355%]

scenario:plugin-cassandra-driver-batch-26

• unstable cpu_user_time [-448.542ms; -186.480ms] or [-19.643%; -8.166%]
• unstable execution_time [-471.498ms; -195.917ms] or [-20.126%; -8.363%]
• unstable instructions [-4.9G instructions; -2.1G instructions] or [-18.389%; -7.880%]
• unstable throughput [+73113.371op/s; +176878.694op/s] or [+8.359%; +20.223%]

scenario:plugin-graphql-long-with-depth-off-26

• unstable max_rss_usage [-49.108MB; -18.565MB] or [-22.459%; -8.490%]

scenario:plugin-graphql-long-with-depth-on-max-20

• unstable cpu_user_time [-1.538s; -0.576s] or [-24.340%; -9.109%]
• unstable execution_time [-1.617s; -0.605s] or [-25.063%; -9.381%]
• unstable max_rss_usage [-29.445MB; -11.130MB] or [-19.095%; -7.218%]
• unstable throughput [+12.635op/s; +33.631op/s] or [+9.391%; +24.997%]

scenario:plugin-graphql-long-with-depth-on-max-26

• unstable max_rss_usage [-53.817MB; -20.311MB] or [-24.699%; -9.322%]

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-04T14:14:42.613305Z	8995319	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[datadog-official]

A synchronous UDP send error can now leave the tracer and stop the application. The LLMObs change also replaces declared tool names with numeric indexes for supported AI SDK 4 array tools.

_{Open Bits AI session}

_{🤖 Datadog Autotest · Commit 8995319 · What is Autotest? · @DataDog review to ask questions · Any feedback? Reach out in #autotest}