diff --git a/doc/plans/2026-10-03-cursor-production-readiness.md b/doc/plans/2026-10-03-cursor-production-readiness.md index 94ed1d57ac6..23ffc0d5a3a 100644 --- a/doc/plans/2026-10-03-cursor-production-readiness.md +++ b/doc/plans/2026-10-03-cursor-production-readiness.md @@ -282,3 +282,7 @@ Review-fix verification passed: 538 controller tests and 38 fixture tests, serve Both affected local repeats passed on runtime c45cf9 with controller/harness df1c65. The first new Daytona denied-write case also passed, including a completed read attestation, exact delivered denial, no automatic resume, continuous absent effects, owned retirement and confirmed sandbox deletion. Remote Stop reached the exact pending-callback cancellation and stale-answer refusal, with a complete remote no-effect watcher and an empty final process journal. Its overall result remains failed: the flow also applied a local API-PID identity check to remote execution, where the controller-command PID legitimately becomes the sandbox runner PID. The repaired flow uses local PID tracking only for local execution. Remote proof still requires the same company/run/lease/sandbox, root boot/start-tick identity, retained descendant journal and continuous watcher. No remote retirement requirement is removed. The failed case and private recovery journal remain retained; only Stop is repeated before continuing the remaining cells. + +The affected remote Stop repeat passed. Remote warm three-turn continuation and normal completion also passed with cleanup. The first remote file case verified exact bytes, successful validation, an accessible registered deliverable and final task/run success. Its overall result remains failed because the automatically generated artifact-preparation comment used the provider-selected file title, which duplicated the completion marker. The fixture now requests the actual filename as its deliverable title and reserves the marker for completion. Exact byte, validation, accessible-output and single-completion checks remain intact; only the file case is repeated before continuing. + +A provider-free normal-install probe also exposed a packaging blocker: the earlier npm consumer had a macOS ARM64 daemon in the universal server package and could not launch it on Linux (exec-format exit 126). Explicit Cursor provisioning itself passed, but that does not certify the installed product. Release packaging must include verified platform-specific daemons for all three promised targets and select the correct bundled daemon without an override. This release fix and the final installed product smoke remain required. diff --git a/tests/runner-e2e/catalog.ts b/tests/runner-e2e/catalog.ts index 31975040020..596ec166dce 100644 --- a/tests/runner-e2e/catalog.ts +++ b/tests/runner-e2e/catalog.ts @@ -1073,6 +1073,7 @@ export const extendedHarnessFileTask: RunnerTaskFixture = { buildPrompt: nonce => [ `Create extended-${nonce}.txt in the current execution workspace with exactly ready-${nonce} followed by a newline.`, `Edit that file to contain exactly verified-${nonce} followed by a newline, then run a command that reads it and fails unless those exact bytes match.`, + `Register the verified file as a deliverable with title extended-${nonce}.txt. Use EXTENDED-FILE-${nonce} only as completion summary and final response, never as a file title or progress message.`, `After successful validation, call paperclip_finish with reportedWorkDisposition done, summary EXTENDED-FILE-${nonce}, the current completion contract revision and satisfied objective criterion, no remaining work, and the actual validation command and result.`, `Wait for paperclip_finish to succeed, then emit exactly EXTENDED-FILE-${nonce} as your final response. Do not create unrelated files or work.`, ].join("\n"),