Codecov Report

❌ Patch coverage is 94.11765% with 4 lines in your changes missing coverage. Please review.
✅ Project coverage is 89.74%. Comparing base (8d64c6e) to head (bae88c2).
⚠️ Report is 7 commits behind head on master.

@@            Coverage Diff             @@
##           master    #8288      +/-   ##
==========================================
+ Coverage   89.63%   89.74%   +0.11%     
==========================================
  Files         836      836              
  Lines       44477    44536      +59     
  Branches     8188     8214      +26     
==========================================
+ Hits        39866    39970     +104     
+ Misses       4611     4566      -45     

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

Overall package size

Self size: 5.78 MB
Deduped: 6.62 MB
No deduping: 6.62 MB

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

✨ Fix all issues with BitsAI or with Cursor

⚠️ Warnings

🧪 3 Tests failed

Plugin kafkajs with kafkajs >=1.4 (1.4.0) data stream monitoring "before each" hook for "Should set a message payload size when consuming a message" from checkpoints     (Fix with Cursor)

This server does not host this topic-partition

KafkaJSProtocolError: This server does not host this topic-partition
    at createErrorFromCode (versions/kafkajs@1.4.0/node_modules/kafkajs/src/protocol/error.js:472:10)
    at Object.parse (versions/kafkajs@1.4.0/node_modules/kafkajs/src/protocol/requests/metadata/v0/response.js:56:11)
    at Connection.send (versions/kafkajs@1.4.0/node_modules/kafkajs/src/network/connection.js:250:35)
    at process.processTicksAndRejections (node:internal/process/task_queues:104:5)
    at async Broker.metadata (versions/kafkajs@1.4.0/node_modules/kafkajs/src/broker/index.js:144:12)
    at async retryOnLeaderNotAvailable.delay (versions/kafkajs@1.4.0/node_modules/kafkajs/src/admin/index.js:98:55)
    at async callback (versions/kafkajs@1.4.0/node_modules/kafkajs/src/admin/index.js:20:14)
...

Plugin kafkajs with kafkajs >=1.4 (2.2.4) data stream monitoring "before each" hook for "Should set a checkpoint on produce" from checkpoints     (Fix with Cursor)

This server does not host this topic-partition

KafkaJSProtocolError: This server does not host this topic-partition
    at createErrorFromCode (versions/kafkajs@>=1.4/node_modules/kafkajs/src/protocol/error.js:581:10)
    at Object.parse (versions/kafkajs@>=1.4/node_modules/kafkajs/src/protocol/requests/metadata/v0/response.js:55:11)
    at Connection.send (versions/kafkajs@>=1.4/node_modules/kafkajs/src/network/connection.js:433:35)
    at process.processTicksAndRejections (node:internal/process/task_queues:104:5)
    at async [private:Broker:sendRequest] (versions/kafkajs@>=1.4/node_modules/kafkajs/src/broker/index.js:904:14)
    at async Broker.metadata (versions/kafkajs@>=1.4/node_modules/kafkajs/src/broker/index.js:177:12)
    at async retryOnLeaderNotAvailable.delay (versions/kafkajs@>=1.4/node_modules/kafkajs/src/admin/index.js:175:55)
...

Plugin kafkajs with kafkajs >=1.4 (1.4.0) without configuration "before each" hook for "should be instrumented" from producer     (Fix with Cursor)

This server does not host this topic-partition

KafkaJSProtocolError: This server does not host this topic-partition
    at createErrorFromCode (versions/kafkajs@1.4.0/node_modules/kafkajs/src/protocol/error.js:472:10)
    at Object.parse (versions/kafkajs@1.4.0/node_modules/kafkajs/src/protocol/requests/metadata/v0/response.js:56:11)
    at Connection.send (versions/kafkajs@1.4.0/node_modules/kafkajs/src/network/connection.js:250:35)
    at process.processTicksAndRejections (node:internal/process/task_queues:104:5)
    at async Broker.metadata (versions/kafkajs@1.4.0/node_modules/kafkajs/src/broker/index.js:144:12)
    at async retryOnLeaderNotAvailable.delay (versions/kafkajs@1.4.0/node_modules/kafkajs/src/admin/index.js:98:55)
    at async callback (versions/kafkajs@1.4.0/node_modules/kafkajs/src/admin/index.js:20:14)
...

ℹ️ Info

🎯 Code Coverage (details)
• Patch Coverage: 74.60%
• Overall Coverage: 86.58% (-0.02%)

Useful? React with 👍 / 👎

Benchmarks

Benchmark execution time: 2026-05-07 18:42:40

Comparing candidate commit bae88c2 in PR branch juan-fernandez/efd-duration-retries-vitest with baseline commit 8d64c6e in branch master.

Found 0 performance improvements and 0 performance regressions! Performance is the same for 1749 metrics, 95 unstable metrics.

💡 Codex Review

Avoid choosing Cucumber retries from coordinator timing

When Cucumber runs in parallel, this code executes in the coordinator and measures the time between receiving testCaseStarted and testCaseFinished messages, not the worker's actual test duration. The worker independently chooses how many EFD retries to run from its local duration in getWrappedRunTestCase, so for tests close to a bucket boundary (for example a 4.99s worker execution with {'5s': 2, '10s': 0} plus IPC/scheduling delay) the coordinator can choose 0 retries and mark the suite finished after the first result while the worker is still emitting retry results. This can produce premature or incorrect suite status in parallel Cucumber runs; the coordinator should use the worker's chosen retry count or avoid recomputing bucket decisions from IPC timing.