Trim RELEASE.md to only repo-specific content

[jnasbyupgrade]

Summary

• The org has consolidated the generic pgxntool release process (versioning, make tag/make dist, the stable pseudo-version dance, manual PGXN upload) into a shared ../ai/RELEASE.md.
• This trims cat_tools' local RELEASE.md down to only what's genuinely repo-specific, per that doc's policy of not keeping a local copy just for symmetry or generic steps.

Kept (genuinely repo-specific):

• The 0.2.3 catalog-view repair / pg_upgrade caveat — real, permanent, user-facing release-notes guidance not covered anywhere else.
• The pre-0.2.2 tagging gap (0.1.x/0.2.0/0.2.1 were never tagged) — this modifies how the shared doc's step 1 drift-check applies to this repo's earliest versions.

Dropped (now generic, covered by ../ai/RELEASE.md):

• All the numbered release steps (safety check mechanics, pre-release checks, version/changelog bump steps, verify, tag/dist, return-to-stable).
• The PGXN_REMOTE-for-forks gotcha — this happened here first but is now documented generically in the shared doc.
• The note that the stable pseudo-version machinery landed shortly after 0.2.3 — verified true (0.2.3 is the last release cut before it landed), but it's historical trivia with no effect on how a future release is actually cut, so it didn't meet the bar to keep.

Dependency

This PR links to ../ai/RELEASE.md, which is not yet merged — depends on Postgres-Extensions/ai#8 landing first.

Test plan

• Confirm CI is green (docs-only change; only the lightweight changes/all-checks-passed jobs should run).
• Confirm the ../ai/RELEASE.md link resolves once Add shared RELEASE.md for PGXN-extension repos ai#8 merges.

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: f509d7ec-4aaa-4f0f-9259-5908f1cdf3b0

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Claude finished @jnasbyupgrade's task in 2m 1s —— View job

---

Code Review: Trim RELEASE.md to only repo-specific content

Tasks

• Gather context (PR diff, description, repo CLAUDE.md conventions)
• Read current RELEASE.md diff in full
• Verify claims in PR description (0.2.3 catalog-view repair, pre-0.2.2 tagging gap, dependency on ai#8)
• Check for dangling references / broken links after trim
• Post review feedback

Summary

This is a clean, well-scoped docs-only trim. I read the full diff (RELEASE.md +22/-143) and checked it against the repo's actual conventions.

Verified correct:

• The two retained sections carry over factual content that was already present (and presumably vetted) in the prior version of RELEASE.md — the pre-0.2.2 tagging gap (0.1.0/0.1.3/0.1.4/0.1.5/0.2.0/0.2.1 never tagged) and the 0.2.3 pg_upgrade catalog-view caveat — so no new unverified claims are being introduced.
• The cross-reference to CLAUDE.md's "SQL file conventions" section is accurate — that section exists and documents the plain-.sql/no-.sql.in exception for the pre-0.2.0 files, matching the new text's parenthetical.
• The README.asc "Updating the extension" section referenced in the pg_upgrade caveat exists.
• The PGXN_REMOTE-for-forks gotcha and the numbered procedural steps are legitimately generic (versioning, make tag/make dist, stable pseudo-version mechanics) — none of it is cat_tools-specific, consistent with the stated policy of not keeping local copies of generic pgxntool process for symmetry.

Things worth confirming before merge (not blockers on this repo's side):

1. This PR's correctness is contingent on Postgres-Extensions/ai#8 (../ai/RELEASE.md) actually landing with equivalent coverage of the dropped steps — that file isn't present in this checkout so I can't verify its content directly, but the PR body already discloses and tracks this dependency correctly.
2. Until ai#8 merges, the [../ai/RELEASE.md](../ai/RELEASE.md) link at the top of the file is dangling. That's expected/acceptable for a stacked PR, but worth double-checking the merge order (this PR should land after, or at least not be the only source of truth, until the target exists) so nobody following the link hits a 404 in the interim.

No code/build/test surface is touched (docs-only), so no functional risk. I did not run CI myself, but per this repo's CLAUDE.md, a doc-only change should only trigger the lightweight changes/all-checks-passed jobs — worth confirming that's what actually ran.

Nice, tight trim — no unnecessary abstractions added, and the two kept sections are genuinely repo-specific with clear justification.