Skip to content

ci: prove signed SmolLM2 Get and Assistant reply in fresh Homes - #153

Merged
andersalm merged 51 commits into
developfrom
feat/0.7.2-ci-core-journeys
Oct 9, 2026
Merged

andersalm merged 51 commits into
developfrom
feat/0.7.2-ci-core-journeys

Conversation

@andersalm

@andersalm andersalm commented Oct 1, 2026 •

Copy link
Copy Markdown
Contributor

CI requires a fresh Home to Get pinned, signed SmolLM2 through Marketplace and display a retained local reply through the normal Assistant picker and Send controls. A separate fresh Home proves Get, explicit engine recovery, Retry and usable Home with the optional engine physically absent.

The existing source-Home jobs use verified licensed engine and Kubo prerequisites. Mac repeats three independent Home states and engine processes. The receipts and visible summaries bind the tested commit/tree, installed Runtime, provider, engine and model; desktop/phone media, disk reserve and owned-process cleanup form part of acceptance. The completed-presentation wait prevents an empty reply from passing after native generation finishes.

Runtime derives fixed host/source prerequisite reasons from admitted content and verified readiness. Model-provider limits actual CPU worker arguments to the smaller of the saved Runtime ceiling and available host CPUs, with one worker when the host query fails. Saved profiles, GPU policy, confinement, cancellation and the existing 120-second deadline retain their values.

The fixture uses a disposable catalogue signing key and locally held Kubo content. Its short reply proves execution and display. The separate-holder, production catalogue, larger-model quality and physical Jetson journeys retain their owning acceptance gates. The existing signed Mac update/Undo journey remains required.

The successful hosted model-proof steps add 9m03s on Mac, 2m26s on Linux ARM and 2m57s on Linux x86 to the existing jobs. Fixture preparation, the installed journey and its summary account for 9m02s, 2m01s and 2m33s respectively. These figures measure the newly added steps. Engine preparation runs inside the existing source setup; a paired full-job comparison remains unmeasured.

Current validation, exact-candidate reviews, installed evidence, successful CI cost and the required irzhywau approval are recorded in #150. PR #208 is merged and its canonical model-contract parity is integrated.

Refs #150, #208.

@andersalm

Copy link
Copy Markdown
Contributor Author

Measured added CI time for candidate 8837e66a in completed run 36856389961: prepare/journey time was ARM64 44/66 seconds, Mac 17/65 seconds, and x86-64 26/64 seconds.
The added block, including Node handoffs, summaries and artifact upload, totaled 119/87/92 seconds respectively: 298 seconds, or 4.97 runner-minutes. The observed critical x86-64 branch contains 92 seconds of added steps.
These are partial failure measurements: all three prepare steps passed, then operator serve returned a bearer-token rejection. Packaging was skipped, so a successful baseline and full critical-path increment remain unavailable.
Published correction 0738258f starts Home gateway and checks its HTTP response. Before publication, a fresh isolated fixture passed HTTP200, enrollment and signed Home using the accepted #154 installed binary; all owned processes and copied fixture data are cleaned up.
The PR description carries the timing table. PR153 stays draft for exact installed replies/screenshots, successful timing, PR133 merge, PR155 integration and required review.

@andersalm

Copy link
Copy Markdown
Contributor Author

Completed CI 36865084148 at 7429962a measured added steps: x86 prepare 44s / failed journey 120s / total 166s; ARM 209s / 121s / 348s; Mac 19s / 120s / 144s. The added steps used 658s (10.97 runner-minutes). The observed critical ARM branch includes 348s of added steps; skipped packaging leaves the successful baseline and full critical-path increment open. These failed-run times are separate from installed acceptance. Candidate 7022fe93 checks gateway /healthz; its isolated installed route proof and four basic gates pass.

@andersalm

Copy link
Copy Markdown
Contributor Author

Independent agent source/security review: round 2 PASS for the Mac runner headroom and safe failure-receipt correction at fc5cdd97b1b552378483fd5d8b66c09991901395. The hosted-runner guard, physical application scope, refreshed Xcode selections, directory identities and final numeric space checks address all three first-round findings. Basic gates, release-policy tests and narrow execution checks pass. The original Mac HTTP409 cause remains unproven; exact installed CI is pending. PR153 stays draft for PR133 merge, reviewed PR155 integration and required human review.

@andersalm andersalm mentioned this pull request Oct 1, 2026
10 tasks
@andersalm andersalm changed the title ci: prove installed Home and pinned model journeys on three platforms ci: prove signed SmolLM2 Get and Assistant reply in fresh Homes Oct 8, 2026
@andersalm

Copy link
Copy Markdown
Contributor Author

Independent development review — VERDICT: PASS; head 064c572, tree 64b030052a6f6aa15e4376e3da1845e4a96eb480.
The independent reviewer covered the full refresh and appended fixes; reviewed ancestry and unchanged paths are verified. Author checks include 2,803 Runtime + 51 model tests and the three-Home installed Mac reply proof.
Current proof, finding dispositions and remaining CI/human gates: #150.

@andersalm

Copy link
Copy Markdown
Contributor Author

Independent adversarial review (Claude Opus via elastos-pr-review) — VERDICT: PASS; head 064c572, tree 64b030052a6f6aa15e4376e3da1845e4a96eb480.
Full current source review found no blockers; diagnostics, signed disposable catalogue, failure privacy and owned-process checks pass review. CI/media/added-time and human approval remain separate acceptance gates.
Nonblocking dispositions and exact installed proof: #150.

@andersalm
andersalm marked this pull request as ready for review October 8, 2026 03:53
@andersalm
andersalm requested a review from irzhywau October 8, 2026 03:53
@andersalm

Copy link
Copy Markdown
Contributor Author

Independent development review: PASS. The current Local AI source and develop integration preserve the reviewed protections; blocking findings: none.
Exact candidate, installed proof and remaining CI/human gates: #150.

@andersalm

Copy link
Copy Markdown
Contributor Author

Independent adversarial review (Claude Opus via elastos-pr-review): VERDICT: PASS.
Blocking findings: none. Source review covers the current published candidate.
Exact candidate, review scope and note dispositions: #150.

@andersalm

Copy link
Copy Markdown
Contributor Author

Operator owns this PR now (8 Oct). Its commits go into train/alpha11 (develop merge 50758f8, ci.yml conflict resolved keeping the journey steps); this PR closes as merged when that train merges. Please don't push here.

@andersalm

Copy link
Copy Markdown
Contributor Author

Carried into #306 (refreshed on develop; merges with the train).

@andersalm
andersalm merged commit 50758f8 into develop Oct 9, 2026
22 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant