Skip to content

Merge upstream panaversity/ksor 0.0.61 into the fork - #17

Draft
salmanparacha wants to merge 7 commits into
mainfrom
chore/sync-upstream-0.0.61
Draft

salmanparacha wants to merge 7 commits into
mainfrom
chore/sync-upstream-0.0.61

Conversation

@salmanparacha

Copy link
Copy Markdown
Owner

Problem

The fork was 4 commits behind panaversity/ksor main (the 0.0.61 npm release).

Solution

Merge upstream/main:

Upstream panaversity#277 fixes the same defect as fork commit 88b1eca. This PR takes upstream's version and removes the fork's, for two reasons. The fork put the override in the shared scaffold template, so the pnpm scaffold carried it too. Upstream's test forbids that, because pnpm's committed lockfile already holds 2.1.2. The fork's duplicate test goes with it.

The merge conflicts were all version or release text:

  • the fork version stays 0.0.61-salman.6;
  • the CHANGELOG keeps the fork entries above upstream's 0.0.61 entry;
  • docs/status.md now names 0.0.61 as the npm line.

Behavior

Nothing changes in the Bedrock/Titan, SigV4, AgentCore credential, region or rollback code. HealthLake stays pinned to v0.0.61-salman.6 until the artifact and HealthLake acceptance steps in docs/fork-maintenance.md (steps 5–7) pass.

Verification

Run locally:

Check Result
install, build, typecheck pass
guard (12 rules) ok
check:corpus ok
unit tests 1775 passed
init, init-manager, site-staging integration suites 96 passed

CI is authoritative for the rest.

🤖 Generated with Claude Code

ziaukhan and others added 7 commits September 21, 2026 18:44
The dev watcher's refreshStage wrote new and edited files into
.staged-knowledge but never removed one, so a document deleted or moved
while `pnpm dev` ran kept answering 200 at its old url and stayed in the
sidebar, and a moved one was listed twice, until a restart.

refreshStage now removes every staged file its plan no longer holds, and
every folder that leaves empty, through the walk pruneSims already had,
extracted as pruneExcept. The 2026-08-18 reason for holding removals back
(a deleted staged file took the dev server down) no longer holds on
fumadocs-mdx 15.4.0 and Next 16.3.3; what remains is a race that clears
itself, recorded in the comment.

Closes panaversity#274.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…naversity#277)

mdast-util-to-markdown 2.1.3 (2026-09-27) writes bold and italic only through
a handler's `attention`. The MDX stringifier of fumadocs-core 16.15.4, which
the site pins, wraps every handler without that property, so the site build
recursed until the stack overflowed (fuma-nama/fumadocs#3604, fixed in
fumadocs-core 16.15.15). The pnpm scaffold kept building because its committed
lockfile holds 2.1.2. The npm and bun scaffolds ship no lockfile, so a fresh
install took 2.1.3, and both manager walks have failed on main since then.

The npm and bun manifests now carry `overrides` pinning 2.1.2, and both
managers honour it. JSON cannot hold a comment, so the README explains the
override to the adopter. The integration test asserts the bound on the emitted
bytes, and checks that the pnpm lockfile holds the same version. It also fails
once the site moves to a fumadocs-core that has the fix, so that the bound
and its README paragraph are removed in that same change.

Closes panaversity#276.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Brings in upstream panaversity#275 (dev stage drops deleted/moved documents), panaversity#277
(hold mdast-util-to-markdown at 2.1.2 in npm/bun scaffolds) and the 0.0.61
version bump.

Upstream panaversity#277 fixes the same defect as fork 88b1eca. Upstream's version is
taken and the fork's is removed: the fork put the override in the shared
scaffold template, so the pnpm scaffold carried it too, which upstream's
test forbids (pnpm's committed lockfile already holds 2.1.2). The fork's
duplicate test in init.integration.test.ts is removed with it.

Conflicts were version/release text only: the fork version stays
0.0.61-salman.6, the changelog keeps the fork entries above upstream's
0.0.61 entry, and status.md now names 0.0.61 as the npm line.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A patch bump from 0.0.61-salman.6 resolves to 0.0.61, the version
Panaversity published to npm. The upstream changes are already in the
CHANGELOG under 0.0.61; fork releases are cut by hand as 0.0.61-salman.N.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
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.

3 participants