chore: bump to 0.0.36 - #108
Conversation
jwfing
left a comment
There was a problem hiding this comment.
Review: chore: bump to 0.0.36
Summary: A clean, correct version bump from 0.0.35 → 0.0.36 across package.json and package-lock.json; the only substantive note is that the PR description understates what this release actually ships.
Requirements context
No docs/superpowers/ or docs/specs/ content exists in this repo, so no matching spec/plan was found — assessed against the PR body, the git history since v0.0.35, and the repo's version-release conventions. This PR is the "bump the version field before tagging" step that version-release expects to be merged to main before a release is cut.
Verification performed
- Version consistency:
package.json(package.json:3) andpackage-lock.json(both the rootversionand thepackages.""entry) are all0.0.36. No source file hardcodes a version — the CLI resolves it dynamically at runtime frompackage.json(src/index.ts:55,src/commands/feedback.ts:75,src/observe/install.ts:11), so no stale constant is left behind. The bump is complete. - Tag idempotency: existing tags stop at
v0.0.35; nov0.0.36tag/release exists, so the follow-up tag release is safe (avoids the irreversible npm-publish collision theversion-releaseskill warns about). - Feature actually present: #106's
feat(deploy)commit (fcd8678) is onmainand reachable from head, so the 0.0.36 release genuinely contains the localdocker buildfallback it references.
Findings
Critical
(none) — the version bump is correct, consistent, and collision-free; nothing blocks merge.
Suggestion
-
Functionality / accuracy of release scope — The body states "No other changes since 0.0.35" and cubic's summary frames this purely as the #106 release. That's not accurate:
git log v0.0.35..HEADshows this release bundles five feature PRs, not one:- #100 — MySQL/MongoDB service commands
- #103 — canonical secret-binding commands
- #104 —
insta buildpre-push verification - #105 —
insta compute exec - #106 — local source deploys (
docker buildfallback)
Blast radius is low because the actual GitHub release notes are auto-generated at tag time (
generate_release_notes: true), so the published changelog will be correct regardless. But the PR description is what a human reviewer reads to decide whether the release is safe to cut, and #103/#105/#100 are non-trivial user-facing surface. Recommend updating the description to list the full set of PRs shipping in 0.0.36 so the reviewer isn't approving a narrower change than what actually ships.
Information
- Commit-message convention — This bump commit is
chore: bump to 0.0.36, whereas prior release bumps consistently usechore(release): 0.0.XX (#NN)(97b0f4e,2d49db0,9d4b3c9, …). Cosmetic only; harmless, but aligning keepsgit log --grepon release commits uniform.
Software engineering / Security / Performance
- Software engineering: No tests are expected or needed for a pure version-field bump; this matches every prior release PR in the repo.
- Security: No security-relevant changes — no code, dependency, or lockfile-dependency changes (only the two
versionstrings in the lockfile moved). - Performance: No performance-relevant changes.
Verdict
approved (informational — no Critical findings; posted as a comment, human approval via the normal flow). The bump is correct and ready to tag; the only ask is a description update to reflect the full 0.0.36 scope, which is non-blocking.
Release cut for #106 —
insta deploy <dir>now works against a local insta-oss daemon (501 from the token mint → localdocker build). No other changes since 0.0.35.🤖 Generated with Claude Code
Summary by cubic
Bumps
instato 0.0.36 to support local source deploys againstinsta-oss. Previouslyinsta deploy <dir>attempted a token mint and failed with 501; now local deploys use a localdocker build, meeting the requirement from issue #106.package.jsonandpackage-lock.jsonversion fields changed; no code diffs.insta deploy <dir>with a runninginsta-ossdaemon to confirm local build path.Written for commit c3c6aa4. Summary will update on new commits.