Skip to content

chore(release): prepare 0.8.5 - #224

Merged
pierreraby merged 1 commit into
mainfrom
release/0.8.5
Nov 17, 2026
Merged

pierreraby merged 1 commit into
mainfrom
release/0.8.5

Conversation

@pierreraby

Copy link
Copy Markdown
Collaborator

Summary

Prepare stable patch 0.8.5 from merged main 7f96417:

Validation

  • Full npm test PASS with both hosts required (PI_LOCAL_REQUIRED=1 OMP_COMPAT_REQUIRED=1, COMMANDCODE_TRANSPORT unset). Local hosts: Pi 1.5.2, OMP 18.22.0, Bun 1.4.4, Node 26.8.1. Host integration uses loopback mocks, not paid or live inference.
  • Entrypoint typecheck against installed Pi 1.5.2 declarations PASS.
  • Release validation contract PASS for v0.8.5/latest with the dated changelog and contributor section; npm test includes the release-rule tests.
  • Format, git diff --check, actionlint and npm audit (0 vulnerabilities) PASS.
  • npm pack --dry-run: 32 files, about 86.0 kB. No packed-package installation performed locally; package and registry checks remain part of the tag-triggered release workflow.
  • Latest CLI catalog check PASS (command-code@1.89.2 current: 80 image, 90 reasoning, 74 effort, 10 output-limit entries); public live pricing check PASS for all 101 models, no unknown prices or unreviewed expired promotions.
  • Preflight: npm latest=0.8.4, next=0.7.3-next.0; latest GitHub Release v0.8.4; npm 0.8.5 E404 and remote v0.8.5 tag absent; no active release run.
  • No migration required; provider ID, credentials namespace, endpoint routing and model output limits are unchanged. Known follow-ups are tracked separately: Bug: No thinking rendering on zai-org/GLM-5.3 #221 (GLM-5.3 thinking), three pre-existing workflow analyzer warnings, and unpinned @latest CI hosts.

Current maintenance is being handled solo. The pricing snapshot received independent review before its merge. Auxiliary workflow-hardening findings concern unchanged baseline workflows and remain separately documented, not suppressed by this PR.

Authorization boundary

This PR is preparation only. Do not merge or create/push v0.8.5 without the operator's explicit authorization. Merging alone does not publish; publication must use the RELEASE.md tag-triggered OIDC workflow. No production installation update is included.

@github-actions

Copy link
Copy Markdown
Contributor

Dependency Review

✅ No vulnerabilities or license issues or OpenSSF Scorecard issues found.

Scanned Files

None

@github-actions

Copy link
Copy Markdown
Contributor

Memory benchmark

Compared base 7f96417 with PR head 53c6cb4 on the same GitHub-hosted darwin runner. Lower values are better.

Metric Base PR PR − Base Change
Stable RSS 107.4 MiB ± 0.3 MiB 106.2 MiB ± 1.3 MiB -1.1 MiB ± 1.4 MiB -1.1%
Physical footprint 61.2 MiB ± 0.6 MiB 60.3 MiB ± 0.5 MiB -1.3 MiB ± 0.3 MiB -2.2%
Physical peak 73.5 MiB ± 0.6 MiB 72.0 MiB ± 1.2 MiB -1.4 MiB ± 2.0 MiB -1.9%

Interpretation: Differences larger than their measured MAD: Physical footprint -1.3 MiB ± 0.3 MiB.

Extension overhead above pi baseline

Metric Base overhead PR overhead Difference
Stable RSS +4.6 MiB ± 1.4 MiB +3.5 MiB ± 1.5 MiB -1.1 MiB ± 1.4 MiB
Physical footprint +3.0 MiB ± 0.5 MiB +2.0 MiB ± 0.4 MiB -1.3 MiB ± 0.3 MiB
Physical peak +2.8 MiB ± 0.5 MiB +1.2 MiB ± 1.4 MiB -1.4 MiB ± 2.0 MiB

Values are medians of 6 alternating, paired runs. The value after ± is the median absolute deviation (MAD). Each process was sampled 12 times after a 2500 ms warm-up.

Environment: pi 0.84.4, Bun 1.4.4, darwin arm64. RSS comes from ps; physical footprint and peak come from macOS footprint.

This is a comparative signal, not a pass/fail threshold. GitHub-hosted runner noise can affect absolute values.

@pierreraby
pierreraby merged commit b3555ad into main Nov 17, 2026
15 checks passed
@pierreraby
pierreraby deleted the release/0.8.5 branch November 17, 2026 10:34
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