Commit graph

10 commits

Author SHA1 Message Date
dan-claude-bot
b6205d81c5 feat(release): CHANGELOG.md and release.yml — a tag becomes the release it claims (#32)
The PR-then-tag half of box#83's flow, rig's side. CHANGELOG.md starts
with an Unreleased section (feature PRs land their entry as part of the
PR — box's convention, now written into CONTRIBUTING alongside the
release ritual). On a tag push, release.yml asserts the bare tag equals
the tree's own VERSION — a mismatch fails loudly and creates nothing —
then creates the GitHub release with that version's changelog section as
the body, extracted by changelog_section in .github/scripts/release-lib.sh:
one function, sourced by the workflow and driven by test/release.sh
against fixtures and the shipped CHANGELOG.md itself. No assets — for a
pure-bash tree, the tag's source tarball IS the package.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-18 20:57:13 +00:00
dan-claude-bot
ec7096c08e fix: base the human auto-request on THIS handoff, not review history
codex's late #85/#98 round-3 finding, valid post-merge: the needs-human
auto-request fired only when the human had NEVER reviewed, so any earlier
human comment or stale approval left a fully-approved PR labeled
needs-human with nobody actually requested — a wedged handoff.
human_request_needed() now asks whether a fresh head-current human review
is missing (live request or head-current approval → nothing to ask;
anything else → request). Five new fixtures cover the wedge, the stale
approval, the satisfied handoff, and request suppression (19 total).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-18 20:16:57 +00:00
dan-claude-bot
03e5deca56 refactor: the author escalates — drop the agreement-body heuristic
Maintainer direction: body-parsing agreement was a guess, and the machine
must not guess. COMMENTED is now unconditionally a non-verdict; the judgment
that a comment-only reviewer's round passed belongs to the PR AUTHOR, who
escalates by requesting the human's review — an explicit request is a fact,
and it is the machine's top-precedence input. Auto-request survives only for
the no-judgment case: three formal head-current approvals. CONTRIBUTING and
LABELS.md state the handoff; fixtures updated (14 transitions, including
author-escalation and the three-formal-approvals path).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-18 19:51:53 +00:00
dan-claude-bot
27fb934780 fix: verdict contract, head-bound approvals, serialized reconcile — and a testable state machine
Round-1 blockers, all three reviewers concurring:
- COMMENTED agreement now counts: agreement_signal recognizes the live bots'
  durable markers (Verdict: Approve / I agree with everything / leading ) —
  the gate to needs-human can actually close. Formal verdicts remain the
  contract (CONTRIBUTING), this is the documented transitional workaround.
- Every counting verdict is bound to the head SHA; a stale approval parks the
  PR in addressing (agent owes re-request) instead of promoting unreviewed
  code. CHANGES_REQUESTED blocks at any head, per GitHub's own semantic.
- reconcile serializes under ONE job-level concurrency group; scope stays
  per-PR. No more cron-vs-event race on the request-the-human-once guard.
- Sweep resilience: per-PR subshell (one failure logs and continues), label
  edits warn instead of wedging; the self-heal claim now matches reality
  (dispatch-only bootstrap).
- The state machine is extracted pure (globals in, state out) and sourceable:
  test/labels-reconcile.sh proves 14 fixture transitions — comment-only
  agreement, stale approval, comment-without-verdict, human precedence and
  human-block — wired into CI.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-18 19:31:02 +00:00
dan-claude-bot
7109bb39d8 feat: label automation — the state reconciler, path-scoped labeler, and CONTRIBUTING
The machinery LABELS.md promised. labels.yml runs the reconciler on a
15-minute cron plus PR events (pull_request_target — every PR here is from a
fork, where pull_request gets a read-only token; no PR code is ever checked
out). The script derives each open PR's state:* from GitHub's own facts and
converges labels statelessly; stale is judged from real activity (commits,
comments, reviews), never label churn, so the sweep cannot un-stale its own
mark. actions/labeler applies scope:* from changed paths. CONTRIBUTING.md is
the guideline: the PR loop, and who sets which labels. Rehearsed with
DRY_RUN=1 against the live repo; shellcheck-clean.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-18 18:31:53 +00:00
Claude
9551ad482f test(db): add real dump/restore round-trip probe + CI job + manual proof docs
The db PR only unit-tested arg parsing; this adds executable proof that
dump/restore actually works end to end.

- test/db-integration.sh: stands up two throwaway Postgres containers whose
  superusers DIFFER by construction (src_super vs dst_super), seeds a known
  checksummable fixture, runs the real `rig db dump`/`rig db restore`, and
  reads the rows back out — proving both invariants db.sh cares about: the
  code reads the container's OWN $POSTGRES_USER/$POSTGRES_DB (a hardcoded
  `postgres` would break on the non-default source superuser), and
  --no-owner --no-acl makes the dump portable across differing superusers (a
  plain dump would abort under ON_ERROR_STOP=1 on the missing role). Also
  asserts default-outfile naming, restore idempotency (--clean --if-exists),
  and the named-[db] scratch-database path. Skips cleanly (exit 0) when Docker
  is absent/unreachable or root is unobtainable; always cleans up via trap.

- ci.yml: separate `db-integration` job on ubuntu-latest (Docker preinstalled),
  kept apart from the fast shellcheck+cli.sh `check` job so an image pull can't
  slow lint feedback.

- README: "Verifying a dump/restore actually works" — the safe manual
  round-trip against a real Coolify container via a fresh scratch db, echoing
  "a backup you have never read back is not yet a backup."

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-17 15:48:51 +00:00
e2a0951a91 ci: enumerate shellcheck inputs with globstar, not by hand
The hand-listed inputs under-covered silently: commands/*.sh does not glob
into lib/, and the previous fix patched that by adding one more path — which
leaves the same hole open for the next subdirectory. Verified: with a broken
script at commands/deep/nested/bad.sh, the hand-listed invocation still exits 0.

`shopt -s globstar` + bin/* **/*.sh covers every script in the repo, including
the extensionless bin/ entrypoints. The list is printed before it runs, so
coverage is visible in the CI log rather than assumed.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-13 19:59:28 +00:00
9c6c6c9477 fix(ci): shellcheck -x, and lint commands/lib/ at all
The runner-* commands source commands/lib/runner-config.sh. Without -x,
shellcheck will not follow that source even though the files carry a
`source=SCRIPTDIR/...` directive, so it reported SC1091 on all three —
and shellcheck exits non-zero on any finding, info-level included.

commands/*.sh does not glob into lib/, so runner-config.sh was being
linted by nothing at all. List it explicitly.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-13 19:56:33 +00:00
df17851108 refactor: rename deployor to rig; canonical heavy-duty/rig URLs 2026-07-11 08:25:48 +00:00
924090a427 feat: curl installer, shellcheck+tests CI, living README 2026-07-11 08:25:11 +00:00