Repository navigation
🌱 chore: record v5 7ef1e44 as merged into v6 (ancestry only, after the squashed top-up) - #19118
Conversation
…ues (#18913) * 🐛 Contributor lane: grace window before claiming owner/hive-filed issues Fixes #18898 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 Fix CI: register SetContributorGrace as a resync setter, drop the settings allowlist entry the UI already covers Refs #18898 Signed-off-by: Andrew Anderson <andy@clubanderson.com> --------- Signed-off-by: Andrew Anderson <andy@clubanderson.com>
…#18917) * 🐛 Label the age, session and staleness of a replayed hand-off decline (#18916) Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 Fix the hand-off UI harness for the stale-reason label (#18916) Signed-off-by: Andy Anderson <andy@clubanderson.com> --------- Signed-off-by: Andy Anderson <andy@clubanderson.com>
…ropped questions (#18920) * ✨ Betty: meeting-budget meter, 80/90 percent steps, and no silently dropped questions (#18914) Meter per meeting and per hive on the meetings read API and in the dashboard; one 80 percent warn line per meeting; shorter answers and no new research at 90 percent; an exhaustion notice that says what to do; questions dropped for budget are counted, shown and listed in the inbox record. Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 Meetings dashboard test: stub the budget meter helpers the role-race slice now needs (#18914) Signed-off-by: Andrew Anderson <andy@clubanderson.com> --------- Signed-off-by: Andrew Anderson <andy@clubanderson.com>
…ach hive (#18918) * ✨ Does this build contain PR #N: ancestry lookup API on the hub and each hive Adds pkg/buildcontains (yes/no/unknown from git ancestry; an unresolvable commit is unknown, a PR outside the build's history is no), the hub read route GET /api/saas/build-contains and the hive route GET /api/version/contains, with openapi, API-reference and telemetry docs. Part of #18915. Signed-off-by: Andy Anderson <andy@clubanderson.com> * ✅ Allow pkg/dashboard to import the leaf pkg/buildcontains resolver Part of #18915. Signed-off-by: Andy Anderson <andy@clubanderson.com> --------- Signed-off-by: Andy Anderson <andy@clubanderson.com>
Automated pin bump (#8419). Digests recomputed from the downloaded release artifacts by src/scripts/cli-pin-bump.sh: VERSION=1.60.1 SHA256_AMD64=0327b1857f8030343f59b77503e0455629bbf472c2901cb94b88389d6278635a SHA256_ARM64=6a962a0ba55829e00f6f297dff736d53b9bdc8da84f0f49c4768e2557ac39cbe Signed-off-by: hive-release[bot] <hive-release@hive.kubestellar.io> Signed-off-by: clubanderson <407614+clubanderson@users.noreply.github.com> Signed-off-by: clubanderson <andy@clubanderson.com>
Automated pin bump (#8419). Digests recomputed from the downloaded release artifacts by src/scripts/cli-pin-bump.sh: VERSION=2.105.0 SHA256_AMD64=79e7ab5db738c48777b696e374f8b428f0647debfb424f44d2f426370407fe2f SHA256_ARM64=b1278c75107423d818f1937fd92cafc9e5e9b606631568f68d7fb816f5885a02 Signed-off-by: hive-release[bot] <hive-release@hive.kubestellar.io> Signed-off-by: clubanderson <407614+clubanderson@users.noreply.github.com> Signed-off-by: clubanderson <andy@clubanderson.com>
Automated release commit. Compiles changelog.d/ fragments and moves the CHANGELOG.md Unreleased section into a dated v5.319.0 entry. See src/docs/releases.md. Signed-off-by: hive-release-bot <actions@github.com>
…ive version menu (#18925) * ✨ Build visibility UI: contains lookup on the hub hive list and the hive version menu Adds a PR number or commit SHA lookup (yes, no or unknown, with the reason) to the hub dashboard hive list and to the hive's own version menu, and shows the newest image commit and the newest merged commit separately. Regenerates static/index.html with make dashboard-gen. Part of #18915. Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🌱 Keep the hub styling ratchet green: tokens and hidden attribute for the contains lookup Part of #18915. Signed-off-by: Andy Anderson <andy@clubanderson.com> --------- Signed-off-by: Andy Anderson <andy@clubanderson.com>
release: v5.319.0
Automated pin bump (#8419). Digests recomputed from the downloaded release artifacts by src/scripts/cli-pin-bump.sh: VERSION=0.170.0 Signed-off-by: hive-release[bot] <hive-release@hive.kubestellar.io> Signed-off-by: clubanderson <407614+clubanderson@users.noreply.github.com> Signed-off-by: clubanderson <andy@clubanderson.com>
…or dashboard changes (#18921) * ✨ UI render evidence: fixture screenshots and one sticky PR comment for dashboard changes Adds a non-blocking pull_request job that renders the affected dashboard pages with synthetic fixture data at desktop and phone width in light and dark themes (no network, no credentials), a workflow_run sibling that posts or updates one sticky comment and records a neutral-or-success check, a doc on reading the evidence, and tests for the pure helpers. Part of #18915. Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🌱 Re-pin TRUSTED_SCRIPTS_TREE for the render-evidence scripts Part of #18915. Signed-off-by: Andy Anderson <andy@clubanderson.com> --------- Signed-off-by: Andy Anderson <andy@clubanderson.com>
Automated pin bump (#8419). Digests recomputed from the downloaded release artifacts by src/scripts/cli-pin-bump.sh: VERSION=2.1.327 Signed-off-by: hive-release[bot] <hive-release@hive.kubestellar.io> Signed-off-by: clubanderson <407614+clubanderson@users.noreply.github.com> Signed-off-by: clubanderson <andy@clubanderson.com>
…d blocks adoption (#18930) * 🐛 Human claim or assignee releases a contributor lease and blocks contributor adoption Fixes #18899 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 📝 Re-anchor the scale-envelope citation after the client.go insertions Refs #18899 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 Fix test fixtures: no claim comment before the owner acts, no time.Sleep Refs #18899 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 Keep the engine's optional transport hooks in the human-claim adoption test Refs #18899 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 Serve an empty comment list in the verdict-cache test Refs #18899 Signed-off-by: Andrew Anderson <andy@clubanderson.com> --------- Signed-off-by: Andrew Anderson <andy@clubanderson.com>
…urrent head (#18893) * 🐛 Hold decisions on a PR whose CI or review is still running on its current head Fixes #18889 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 mid-flight: read the nested PR's park reason from its listed row Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 mid-flight: mirror new rules in the JS glossary, fix pending-CI test, document skips Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 mid-flight: tidy the skipped-PR word, regenerate the bundle Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 mid-flight: fail closed when the review state cannot be read Signed-off-by: Andrew Anderson <andy@clubanderson.com>
Automated release commit. Compiles changelog.d/ fragments and moves the CHANGELOG.md Unreleased section into a dated v5.320.0 entry. See src/docs/releases.md. Signed-off-by: hive-release-bot <actions@github.com>
release: v5.320.0
… released (#18928) * fix: refuse relay requests once the per-kick key is released Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com> Co-authored-by: clubanderson <407614+clubanderson@users.noreply.github.com> * 🐛 test: drain response body in relay test roundTrip The released-key tunnel test reuses one TLS connection for two requests. roundTrip built a fresh bufio.Reader and never read the body, so the leftover 'ok' body was parsed as part of the next status line (malformed HTTP version "okHTTP/1.1"). Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com>
Automated release commit. Compiles changelog.d/ fragments and moves the CHANGELOG.md Unreleased section into a dated v5.320.1 entry. See src/docs/releases.md. Signed-off-by: hive-release-bot <actions@github.com>
release: v5.320.1
Automated release commit. Compiles changelog.d/ fragments and moves the CHANGELOG.md Unreleased section into a dated v5.320.2 entry. See src/docs/releases.md. Signed-off-by: hive-release-bot <actions@github.com>
release: v5.320.2
…e after a switch (#18912) * 🐛 fix: model/method switch for a stopped or paused agent, and keep the page in place after a switch Fixes the not-found failure for configured agents the runtime does not hold and holds scroll, focus and card order across the re-render. Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 fix: import sort in the agent control handlers Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 fix: keep a second body tag out of the generated dashboard document Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 fix: test a registered agent's model switch without assuming it is launched Signed-off-by: Andrew Anderson <andy@clubanderson.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com>
Automated release commit. Compiles changelog.d/ fragments and moves the CHANGELOG.md Unreleased section into a dated v5.320.3 entry. See src/docs/releases.md. Signed-off-by: hive-release-bot <actions@github.com>
release: v5.320.3
Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com>
Automated release commit. Compiles changelog.d/ fragments and moves the CHANGELOG.md Unreleased section into a dated v5.320.4 entry. See src/docs/releases.md. Signed-off-by: hive-release-bot <actions@github.com>
…sed (#18943) Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com>
release: v5.320.4
…#19041) ValidateLinkURL only parsed strict IP literals, so a malformed all-numeric host reached DNS and loose IPv4 spellings were caught only after resolution. It now calls netguard.HostDenyReason, which the security docs already describe. Signed-off-by: Andrew Anderson <andy@clubanderson.com>
Automated release commit. Compiles changelog.d/ fragments and moves the CHANGELOG.md Unreleased section into a dated v5.323.3 entry. See src/docs/releases.md. Signed-off-by: hive-release-bot <actions@github.com>
…efault off) (#19047) * ✨ PR gate: post the PR-contents check from the hive in shadow mode GitHub blocks pull_request_target workflows by default in public repositories from 2026-12-02, and the required "verify PR contents" check comes from one. This is the first slice of posting that gate from the hive itself as a commit status, shipped default off and shadow-only. - src/pkg/prgate: Go port of the PR Type / PR Issue verdicts, replayed against the same golden corpus as the JS verifier so they cannot drift; a closed four-context allowlist enforced before every write; a gate that pins the event head, re-reads the PR, skips drafts and closed PRs, is idempotent, and records a comparison with the workflow result. - The status write now goes through one shared writer (also used by the prow/lgtm binding, behaviour unchanged). - auto_merge.pr_gate.mode (off, shadow, enforce) with an explicit repos scope, yaml-only. enforce runs as shadow until the grant check lands. - The authenticated pull_request and check_run webhook deliveries feed the gate; with the mode off nothing is read or posted. Refs #19038 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 PR gate: route the dashboard through the github package and update the config golden pkg/dashboard keeps its internal import count by reaching the gate through a facade in pkg/github, and the agent runtime config golden carries the new empty auto_merge.pr_gate block. Refs #19038 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 📖 docs: refresh api-reference citations after the server.go change Refs #19038 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 PR gate: bound the memory of posted statuses The posted-status memory was the one unbounded map in the gate; it now restarts empty at a fixed size like the verdict memory does. Refs #19038 Signed-off-by: Andrew Anderson <andy@clubanderson.com>
release: v5.323.3
…turn a green head red (#19054) * 🐛 merge gate: a blocked pull_request_target startup failure does not turn a green head red When the required context set is known and every required context is green, a zero-job startup_failure run of event pull_request_target is logged instead of failing the merge-request CI verdict. Every other zero-job failure, and any head whose required contexts are unknown, missing or red, stays red. Refs #19038 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 merge gate: fix the duplicated struct field in the target startup change Refs #19038 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * 🐛 merge gate: make the missing-required-context test use branch protection Refs #19038 Signed-off-by: Andrew Anderson <andy@clubanderson.com> --------- Signed-off-by: Andrew Anderson <andy@clubanderson.com>
… fetchers (#19030) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com>
…ld, remote) (#19029) * 🐛 fix: answer human review feedback on the hive's own agent PRs (draft, held, remote) Fixes #19027 Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 fix: mark the needs-human note as solicited for the write-back ratchet Fixes #19027 Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 fix: tag the needs-human note as a notification site and fix the feedback test fixture Fixes #19027 Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 fix: coalesce review feedback bursts, queue mid-run comments, rolling cap, whole-thread briefs Fixes #19027 Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 test: raise reviewfeedback coverage over the package floor Fixes #19027 Signed-off-by: Andy Anderson <andy@clubanderson.com> --------- Signed-off-by: Andy Anderson <andy@clubanderson.com>
Automated release commit. Compiles changelog.d/ fragments and moves the CHANGELOG.md Unreleased section into a dated v5.324.0 entry. See src/docs/releases.md. Signed-off-by: hive-release-bot <actions@github.com>
release: v5.324.0
… failure reason (#19061) * 🐛 fix(contribute): skip opencode --auto when unsupported and show the failure reason Fixes #19060 Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🧪 test(contribute): run the failure-tail test on the interactive path Signed-off-by: Andy Anderson <andy@clubanderson.com> --------- Signed-off-by: Andy Anderson <andy@clubanderson.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> Signed-off-by: scanner <280983584+hivecommons-hive[bot]@users.noreply.github.com>
…lls back to podman (#19077) * 🐛 contributor container: support rootless Docker, fall back to podman on permission failure Fixes #19076 Signed-off-by: Andrew Anderson <andy@clubanderson.com> * ci: regenerate integration candidate workflows for runtime-selection test (#19076) Signed-off-by: Andrew Anderson <andy@clubanderson.com> * test: handle runtime-fallback interpolations in container-exit block harness (#19076) Signed-off-by: Andrew Anderson <andy@clubanderson.com> --------- Signed-off-by: Andrew Anderson <andy@clubanderson.com>
… when the queue is empty (#19053) * 🐛 fix(contribute): pre-accept Claude's bypass consent, never Enter on "No, exit", and say plainly when the queue is empty Fixes #19049 Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 fix(contribute): reset pane capture counter in consent test Refs #19049 Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 fix(contribute): multi-hive aware empty-queue messaging, retry bookkeeping behind HIVE_RELAY_DEBUG Refs #19049 Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 fix(contribute): no_matching_work never claims an empty queue; use the hive's actionable_items count Refs #19049 Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 fix(contribute): correct short-name test expectation Refs #19049 Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 fix(contribute): cover the hive-stated name in hiveShortName Refs #19049 Signed-off-by: Andy Anderson <andy@clubanderson.com> --------- Signed-off-by: Andy Anderson <andy@clubanderson.com>
) * 🔎 Add independent fact-check review pass for analysis and report PRs Opt-in review.analysis_factcheck: a reviewer agent that is not the author re-verifies claims, scores and attributions, checks skew both ways and inspection coverage, checks people against maintainers files, and posts one review with line-anchored Verify comments. The PR stays a draft; rounds are bounded and findings reach the author through a marked path of their own. Fixes #19074 Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 Fix promoted-field literal in fact-check wire test Signed-off-by: Andy Anderson <andy@clubanderson.com> * 🐛 Add analysis_factcheck to the agent runtime config golden Signed-off-by: Andy Anderson <andy@clubanderson.com> --------- Signed-off-by: Andy Anderson <andy@clubanderson.com>
#19100 carried v5 up to 7ef1e44 (#19080) into v6 but landed as a squash (f4f3849), so v6 has the content without the v5 history. This merge records 7ef1e44 as a parent with v6's tree unchanged (git merge -s ours), so the next v5 to v6 top-up starts from the right merge base. Signed-off-by: Andy Anderson <andy@clubanderson.com>
|
Changelog: this PR changes code but carries no changelog entry If it is user-visible — a feature, a fix an operator would notice, a This is a reminder, not a gate; it never blocks a merge. |
|
Hive status — updated in place; edits do not notify. Full history is in the Activity Monitor. @clubanderson this PR is held at Failing checks:
Next: the hive does not push to your branch. Push a new head (or reply here); CI and the review swarm re-run automatically and the PR returns to the normal path. |
|
[APPROVALNOTIFIER] This PR is APPROVED Approval requirements bypassed by manually added approval. This pull-request has been approved by: The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
This PR records v5 7ef1e44 (#19080) as merged into
v6without changing any content.#19100 brought v5 up to 7ef1e44 into v6, but Tide squashed it into f4f3849, which has one parent. v6 has the content but not the v5 history, so the next v5 to v6 top-up would conflict on everything again. This branch is a single
git merge -s ourscommit, f4f3849 + 7ef1e44:git diff origin/v6 HEADis empty: v6's tree is unchanged.git merge-base --is-ancestorpasses for 7ef1e44, d5b0d0f, 0de5fef, de3a4ee, 41e6d19 and e1d4b26.It must land as a merge commit (
tide/merge-method-merge), never a squash or a rebase. A squash would drop the second parent again. v5-sync-policy.md step 3 calls for a real merge commit for sync PRs.Tide cannot land this: its local merge of an
-s ourscommit has nothing to commit. It is merged withgh pr merge --merge, the way #18341 and #18508 were.🤖 Generated with Claude Code