Toolkit main run 5926 succeeded but took 32m30s on the static verifier. The preceding full main run 5830 took 15m59s on an Alkane GARM worker. The static job overlapped SDK board compilation. Its microbenchmark badge reported a 95.5% regression despite only CI files changing. The measured difference is likely runner contention, not a source-code regression.
Change
Return the full main release job to the dedicated alkane-ci build pool. Keep fast PR verification on alkane-verify, and retain all main tests, coverage, benchmarks, badges, Pages trigger, release checks, and cleanup.
Validation
The runner image gate, workflow YAML parse, and whitespace check pass locally. A full main run after merge must verify the actual assigned runner, elapsed time, cache behavior, benchmark baseline, and artifact cleanup. The alkane-ci label is shared by the static and GARM runners, so this change does not guarantee GARM selection; the CI result will determine whether further pool isolation is needed.
## Why
Toolkit main run 5926 succeeded but took 32m30s on the static verifier. The preceding full main run 5830 took 15m59s on an Alkane GARM worker. The static job overlapped SDK board compilation. Its microbenchmark badge reported a 95.5% regression despite only CI files changing. The measured difference is likely runner contention, not a source-code regression.
## Change
Return the full main release job to the dedicated alkane-ci build pool. Keep fast PR verification on alkane-verify, and retain all main tests, coverage, benchmarks, badges, Pages trigger, release checks, and cleanup.
## Validation
The runner image gate, workflow YAML parse, and whitespace check pass locally. A full main run after merge must verify the actual assigned runner, elapsed time, cache behavior, benchmark baseline, and artifact cleanup. The alkane-ci label is shared by the static and GARM runners, so this change does not guarantee GARM selection; the CI result will determine whether further pool isolation is needed.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Why
Toolkit main run 5926 succeeded but took 32m30s on the static verifier. The preceding full main run 5830 took 15m59s on an Alkane GARM worker. The static job overlapped SDK board compilation. Its microbenchmark badge reported a 95.5% regression despite only CI files changing. The measured difference is likely runner contention, not a source-code regression.
Change
Return the full main release job to the dedicated alkane-ci build pool. Keep fast PR verification on alkane-verify, and retain all main tests, coverage, benchmarks, badges, Pages trigger, release checks, and cleanup.
Validation
The runner image gate, workflow YAML parse, and whitespace check pass locally. A full main run after merge must verify the actual assigned runner, elapsed time, cache behavior, benchmark baseline, and artifact cleanup. The alkane-ci label is shared by the static and GARM runners, so this change does not guarantee GARM selection; the CI result will determine whether further pool isolation is needed.