Repository navigation
chore(deps): bump @adcp/sdk to 17.0.0-beta.29 and move the test-agent preview to 3.3-rc.0 - #9072
Merged
Merged
Conversation
… preview to 3.3-rc.0
@adcp/sdk 17.0.0-beta.29 pins AdCP 3.3.0-rc.0, and Python adcp 10.0.0b20
adopts the same bundle. tests/sdk-runtime-compat ties Choose your SDK and
the training agent to the installed SDK, so bump it, and move
TRAINING_AGENT_HOSTED_PREVIEW to { wireVersion: '3.3-rc.0', bundle:
'3.3.0-rc.0' }, the prerelease beta.29 negotiates. The preview-version
and tenant-smoke tests follow, with 3.3-beta.6 as the previous preview.
Docs, replacing the forward-looking rc.0 text from #8883:
- Choose your SDK names 17.0.0-beta.29 and 10.0.0b20, both on 3.3.0-rc.0.
- The prerelease page: 17.0.0-beta.24 to beta.28 and 10.0.0b17 to b19
support beta.6 (b19's wheel bundles 3.3.0-beta.6), beta.29 and b20
support rc.0, and the test-agent preview serves 3.3-rc.0. It also says
what the agent refuses: installment_refs (#8986), and the assignment
operations each route lists (#8997).
- The release notes, Versions, the quickstart, and the creative33
sandbox note name the new versions and pin.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Contributor
There was a problem hiding this comment.
Ladon verdict: Approve
Approve — clean, mechanical version bump with no blocking or medium findings.
Checked:
@adcp/sdk17.0.0-beta.28→beta.29 and hosted test-agent preview3.3-beta.6→3.3-rc.0move consistently across the centralized preview constant (candidate-version.ts), both tenant-smokesupported_versionslists, and the preview-version unit test; refusal/non-match cases correct.- SDK support matrices agree across choose-your-sdk, 3-3-beta, release-notes, and versions docs.
- Immutable
dist/schemas/3.3.0-beta.6/**release artifacts untouched. - Changeset present and correctly typed
patch; no breaking wire change.
High-risk flag is true only because three docs/reference/** files are modified; reviewer found no medium-or-higher concerns on those files, so the modifications are presumed safe (not an escalation trigger per the high-risk heuristic). No gated paths, no no-auto-approve team match, no deletions. The author merge-hold pending Python adcp 10.0.0b20 on PyPI is a process gate, not a code finding.
No findings of any severity → rows 1–8 do not fire → row 9: approve.
This was referenced Nov 17, 2026
bokelley
added a commit
that referenced
this pull request
Nov 18, 2026
… on (#9082) * fix(server): retry a failed docs index and log what /ready is waiting on Production deploys have failed since #9072: in each bluegreen rollout one green machine never passed /ready within Fly's 10-minute wait_timeout and Fly rolled back (deploys of c925886 and 954eb9d). /ready waits on web chat, which waits on Addie's knowledge index, and that left two gaps: - If initializeDocsIndex() threw, knowledge search stayed unready until "a later caller" retried, and after boot nothing calls it again, so one failed attempt stranded the machine. It now retries on its own with backoff (5s, 15s, 30s, then every 60s) until it succeeds. - The failure was logged as { error }, which the logger does not serialize, so the reason never appeared. It is now logged as { err } with memory usage. /ready now logs which checks block it (database, reporting, chat) with the database and reporting errors and, for chat, which part of its knowledge index is missing; once per change and at most every 60s while unchanged. Fly does not keep probe bodies, so this is how the next stuck rollout names its blocker. The 3.3-rc docs snapshot (#9071, #9072) took the docs index from 5 to 6 versions: locally, under the web process's 2 GB heap limit, peak heap during indexing went from 488 MB to 703 MB and indexing time from 12s to 19s. That is the likely trigger, but the stuck machine's own logs were lost at rollback, so this change makes the next attempt recover or report. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * chore(addie): refresh the tool-surface inventory hash for addie-chat.ts The /ready diagnostics change added webChatReadiness() to server/src/routes/addie-chat.ts, so its source hash in the generated inventory moved. Regenerated with npm run build:addie-tools; npm run test:addie-tools passes. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
bokelley
added a commit
that referenced
this pull request
Nov 18, 2026
…#9088) * ci(deploy): record green machines' logs and events during the rollout Every production deploy since #9072 has failed the same way: one of the two bluegreen green machines never passes /ready, Fly rolls back after the 10-minute wait and deletes it, and its boot output and events go with it. The forensics step only runs after the rollback, so its 500-line log tail has never contained a line from the stuck machine. The Deploy step now, for the whole rollout: - streams `flyctl logs -a adcp-docs` to a file (stopped on every exit path by a trap); and - every 15s records the status (`flyctl machine status -d`), checks and events of each machine that did not exist before the deploy, replacing a snapshot only when a new one succeeds, so the last one taken before a rollback deletes the machine survives. On failure, the forensics step applies the existing secret redaction to all of it and adds fly-logs-deploy.txt (last 20,000 lines), one fly-logs-green-<id>.txt per green machine, green-<id>.json (state, checks, events, including any OOM exit) and green-<id>-status.txt, and lists the green machines in summary.txt. Checked with a fake flyctl: a green machine deleted mid-rollout keeps its last snapshot and boot lines, secret-shaped lines are redacted, and the log stream stops when the step exits. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * ci(deploy): upload only the green machines' redacted log lines Review on #9088: uploading the whole rollout's log stream (up to 20,000 lines, including the serving fleet's request logs) through a best-effort redaction regex into a 30-day artifact widened the exposure surface far beyond what the diagnosis needs. The stream now stays on the runner. On failure the forensics step uploads only each green machine's own lines (redacted, last 3,000 per machine), which are boot output from machines that serve no traffic until they pass /ready, plus their redacted status, checks and events as before. fly-logs-deploy.txt is gone. Re-checked with a fake flyctl: no serving-fleet line reaches the artifact, green machines' lines and a deleted machine's last snapshot survive, and Bearer lines are redacted. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * ci(deploy): keep the deploy's own errexit toggle first in the Deploy step deploy-workflow.test.ts runs the Deploy step's flyctl calls by slicing the script from the first `set +e` to `deploy_exit=$?`. The green-machine watcher added a `set +e` of its own earlier in the script, so the slice started inside the watcher loop and ran it in the foreground forever: Server unit tests (2/4) hung until the 20-minute job timeout on both CI runs of #9088 (the test runs in 64ms on main). The watcher now uses `set +o errexit`, so the slice starts at the deploy's own toggle again. deploy-workflow.test.ts passes (24 tests). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * ci(deploy): fail closed when the pre-deploy machine list is unavailable Review on #9088: green machines are "every machine not listed before the deploy", and `flyctl machines list ... || true` left that list empty on a transient Machines API failure, which would have classified the serving fleet as green and uploaded its request logs on a failed deploy. The list is now retried three times and must be non-empty. Without it, the step warns and captures nothing: no log stream, no watcher, so no machine's lines can reach the artifact. stop_capture tolerates either background job being absent. Simulated with a fake flyctl: with the list available only the green machines are captured and no serving-fleet line reaches the artifact; with the list failing, nothing is captured. deploy-workflow.test.ts passes (24 tests). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * ci(deploy): bound the green-machine watcher so it cannot hang the deploy Review on #9088: the EXIT trap waited for the watcher, whose flyctl calls had no timeout, so a hung Machines API call during a failing deploy (the case this capture exists for) would have held the Deploy step until the job timeout, with no forensics uploaded. Each watcher flyctl call now runs under `timeout 30`, and stop_capture kills the watcher before reaping it. Snapshots are moved into place whole, so the last good one survives the kill. Simulated with a fake flyctl whose `machine status` hangs: the step exits as soon as the deploy work does, and the snapshot taken before the hang is kept. deploy-workflow.test.ts passes (24 tests). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
3.3.0-rc.0is published (tagv3.3.0-rc.0atb5ac0e1bf7). This is the post-cut follow-up to #8883. It names the SDK prereleases that bundle rc.0 and moves the public test agent's preview to rc.0.Changes
@adcp/sdkgoes from 17.0.0-beta.28 to 17.0.0-beta.29, which pins AdCP3.3.0-rc.0and is on npmbeta.tests/sdk-runtime-compat.test.cjsties Choose your SDK and the training agent to the installed SDK, so this bump has to come first.TRAINING_AGENT_HOSTED_PREVIEWmoves to{ wireVersion: '3.3-rc.0', bundle: '3.3.0-rc.0' }, the prerelease beta.29 negotiates. The committeddist/schemas/3.3.0-rc.0bundle is the validation root.training-agent-preview-version.test.tsand the twotenant-smokesupported_versionslists follow, with3.3-beta.6as the previous preview and3.3-rc.1as the not-listed later prerelease.@adcp/sdk17.0.0-beta.29 and Pythonadcp10.0.0b20, both on3.3.0-rc.0.3.3-rc.0;installment_refswithUNSUPPORTED_FEATURE(fix(training-agent): refuse installment_refs with UNSUPPORTED_FEATURE #8986), and thecreative.assignment_operationseach route lists (feat(schema): declare the sync_creatives assignment operations a seller accepts #8997).known-findings.json.rc0-sdk-and-test-agent-preview.md(patch).Python 10.0.0b18 and b19 are on PyPI, published before rc.0. b19's wheel bundles
3.3.0-beta.6, which is the basis for "10.0.0b17 through 10.0.0b19 support beta.6".Checks run
test:unitpassed.precommit:server-unithad one failure in each of three shards:hosted-compliance-echoed-release,matched-pair-ni/exact-inference, andtraining-agent-hosted-grader-credential. The hook's own rerun of those three files passed.tests/sdk-runtime-compat.test.cjs: 9/9 against@adcp/sdk17.0.0-beta.29 (ADCP_VERSION3.3.0-rc.0).training-agent-preview-version.test.tsandserved-release-shape.test.ts: 28/28.tenant-smoke.test.ts: 55/56. The one 15 s timeout, "reads controller-seeded media buys…", passes when run alone.verify-training-agent-runtime-assets:3.3-rc.0resolves todist/schemas/3.3.0-rc.0.test:docs-nav,check:owned-links,lint:schema-links, and the changeset scope check./signals(57 clean against a floor of 45). It was stopped partway through/salesafter 30 minutes, and the branch was pushed withADCP_SKIP_STORYBOARD_MATRIX=1. CI runs the storyboards.🤖 Generated with Claude Code