Graph task: module-sdk-browser-harness-hostdispatch-arena-fail · state: claimed · priority: 2 · components: [module-sdk]
Agent: sdk-browser-fix
Observed 2026-07-29 while running the gauntlet for wasm-plaintext-memory-hygiene. Independent of that task's diff (it touched no SDK source).
cd repos/ancillary-packages/space-data-module-sdk
node --test test/node-host.test.js test/browser-host.test.js test/browser-module-harness.test.js test/testing-harness.test.js
-> 44 tests, 42 pass, 2 FAIL
Failing:
- browser module harness exposes awaited host dispatch alongside module invoke
- browser module harness external arena direct invoke rejects non-shared module memory
That is the gauntlet's REQUIRED module-sdk/sdk-host-surfaces check, so every module-sdk task is currently gated behind it. Local HEAD is 0bd7688 (v0.8.4-87), which is BEHIND origin/main 2a0fdc7 — first question to answer is whether main already fixes these (fast-forward, re-run) or whether they are real regressions; see orbpro-sdk-submodule-pin-divergence.
Also seen in the same run: the gauntlet process died with exit 144 during module-sdk/sdk-full-suite, so that check never reported. Worth confirming whether the runner is hitting a timeout/kill path of its own.
Mirrored from the dev graph (graph/tasks/module-sdk-browser-harness-hostdispatch-arena-fail.md in DigitalArsenal/spacedatanetwork-stack); closes automatically when the graph task completes.
Graph task:
module-sdk-browser-harness-hostdispatch-arena-fail· state: claimed · priority: 2 · components: [module-sdk]Agent: sdk-browser-fix
Observed 2026-07-29 while running the gauntlet for wasm-plaintext-memory-hygiene. Independent of that task's diff (it touched no SDK source).
cd repos/ancillary-packages/space-data-module-sdk
node --test test/node-host.test.js test/browser-host.test.js test/browser-module-harness.test.js test/testing-harness.test.js
-> 44 tests, 42 pass, 2 FAIL
Failing:
That is the gauntlet's REQUIRED module-sdk/sdk-host-surfaces check, so every module-sdk task is currently gated behind it. Local HEAD is 0bd7688 (v0.8.4-87), which is BEHIND origin/main 2a0fdc7 — first question to answer is whether main already fixes these (fast-forward, re-run) or whether they are real regressions; see orbpro-sdk-submodule-pin-divergence.
Also seen in the same run: the gauntlet process died with exit 144 during module-sdk/sdk-full-suite, so that check never reported. Worth confirming whether the runner is hitting a timeout/kill path of its own.
Mirrored from the dev graph (
graph/tasks/module-sdk-browser-harness-hostdispatch-arena-fail.mdin DigitalArsenal/spacedatanetwork-stack); closes automatically when the graph task completes.