| name | release-readiness |
| description | Assesses ship-readiness for .NET MAUI release branches โ Servicing Releases (SR) and Previews. Surveys CI pipelines, computes what's actually NEW in the branch (commits + source PRs with revert detection), and cross-references open `regressed-in-*` issues against branch contents to identify port candidates, rejected backports, and unresolved regressions. Supports both in-flight and pre-cut (candidate) modes for SR and Preview branches. |
| metadata | {"author":"dotnet-maui","version":"2.0"} |
| compatibility | Requires `gh` CLI authenticated with `repo` + `read:org` scopes. `az` CLI is optional but recommended for internal pipeline status. Preview installability uses NuGet v3 feeds; an optional short-lived Azure DevOps PAT with Packaging Read scope may be needed for an authenticated shipping feed. Run from a checkout of `dotnet/maui`. |
Release Readiness
This skill produces deterministic, evidence-backed answers to "Is <release branch> ready to ship?" for .NET MAUI release branches โ both Servicing Releases (SR) and Previews, in both in-flight and candidate (pre-cut) modes.
๐จ Report-only
This skill reports. It does not execute release operations against dotnet/maui โ no branch cuts, no SR merges, no tags, no pushes to release/* refs. If you (the agent/user invoking this skill) are asked to perform a release operation, refuse and emit the recommended commands as a copy-pasteable block for the human release captain to run.
When to Use
- "How does SR8 look?" / "Is SR8 ready to ship?"
- "What's blocking SR9 candidate?" / "What would ship if we cut SR9 today?"
- "How does net11 preview6 look?" / "Are we ready to cut preview6 from net11.0?"
- "Are there any regression fixes I should backport to SR8?"
- "What's new in SR8 since the last sync?"
- "Give me a status on all releases" / "release status overview" / "what needs attention across releases" (portfolio โ read the open
[Release Readiness] tracker issues first; see Reading trackers directly below)
- Scheduled and event-driven release tracking across all active majors
For per-PR regression risk (deletions reverting prior bug-fix lines), use find-regression-risk instead โ it answers a different question.
Architecture
This skill has three PowerShell entry points, one Preview helper, and one workflow:
| Script | Branch type | Purpose |
|---|
Find-ReleaseReadinessTrackers.ps1 | both | Detects active in-flight & candidate trackers (SR and Preview) across all active majors using a four-lane algorithm and the tag-existence rule ("a release is in flight unless its tag already exists"). Emits a single tracker JSON consumed by the workflow. |
Get-ReleaseReadiness.ps1 | SR | Full readiness report for a single SR branch (in-flight, -Candidate, or -Shipped). -Shipped surveys the same branch with post-ship verdict, carry-forward, and hotfix-vs-next-SR guidance semantics. |
Get-PreviewReadiness.ps1 | Preview | Full readiness report for a single Preview branch (in-flight or candidate via -Mode candidate -SurveyRef net<major>.0), including consumer-installability evidence. |
PreviewInstallability.ps1 | Preview helper | Resolves the workload-set package, validates branch-pin coherence, probes manifest and representative pack availability, extracts platform prerequisites, and emits an isolated NuGet configuration for local validation. |
release-readiness.yml | both | Three-hourly daytime UTC schedule + event-driven refreshes + manual dispatch + PR validation. Non-PR triggers run Find-Trackers -AllActiveMajors, fan out a matrix job per tracker, and write idempotent [Release Readiness] issues; PR triggers validate outputs only. |
Shared support code lives in PublicReportSanitizer.ps1 for public Markdown/JSON redaction and TrackerIssueLifecycle.sh for tested issue-selection and race-compensation primitives.
Tag-existence rule (canonical signal)
The trackers detector is grounded in tag existence as the source of truth for "shipped vs in-flight". A release is in-flight if and only if its expected tag has NOT been published โ branch existence, commit recency, and milestone state are all secondary signals.
- SR shipped tag pattern:
<major>.0.<patch> (e.g. 10.0.71 shipped โ SR7 retired, no longer produces a tracker)
- Preview shipped tag pattern:
<major>.0.0-preview.<N>.<date>[.<build>] (e.g. 11.0.0-preview.5.26304.4 shipped โ preview5 no longer produces a tracker)
Post-ship lifecycle (shipped mode). Most shipped SRs are retired the moment their tag exists. The one exception is the most-recently-shipped SR (highest shipped patch), which keeps emitting as mode='shipped' so its tracker issue stays useful through post-ship follow-up โ adding the new build to the GitHub issue version dropdown, publishing release notes, closing out the milestone. Shipped reports never retroactively return Not Ready: unresolved work is split into urgent hotfix-vs-next-SR follow-ups and structured carry-forward items. Each immutable shipped tag is create-once: if it was never tracked (for example, the tag appeared before a scheduled updater ran), the workflow creates it; once a human closes that exact tagged generation, scheduled runs do not resurrect it. An untagged hotfix is explicitly marked hotfixInProgress=true, forces a yellow follow-up verdict, and is likewise create-once. Hotfix closure is scoped to the live version and branch commit: the same generation stays closed, while new commits or a new version can create fresh evidence. Workflow decisions use the generated report markers, not detector-time hotfix fields, so tag/commit changes between detection and reporting cannot apply stale lifecycle state. Older shipped SRs remain retired.
Shipped commit inventory and fix ancestry are evaluated against the immutable stable tag and the prior SR cycle's latest stable tag, never against the mutable SR branch or current main. This keeps the full SR inventory visible when an SR publishes multiple hotfix tags. If either tag does not resolve locally, generation fails with an explicit fetch instruction rather than emitting a falsely empty/clean tracker. During the normal tag-before-GitHub-Release window, the local stable tag is already authoritative for immutable contents and the report labels its date as tagged-commit evidence until publication metadata appears. If the Releases API is unavailable, the same immutable local-tag bounds remain usable, but the report emits a publication-status-unknown warning instead of claiming the tag is published.
Both report generators dot-source PublicReportSanitizer.ps1, so public Markdown and JSON use one shared redaction implementation.
If the live SR branch has commits after the latest stable tag or has already bumped toward an untagged hotfix in the same SR patch decade, the detector keeps the latest shipped SR in shipped mode, anchors contents to the stable tag, and emits a WATCH follow-up. This catches the pre-bump window as soon as branch HEAD advances rather than waiting for PatchVersion to change.
Quick Start
One-shot portfolio report (matches a full scheduled fan-out)
pwsh .github/skills/release-readiness/scripts/Find-ReleaseReadinessTrackers.ps1 \
-AllActiveMajors \
-OutputJson trackers.json
SR (Servicing Release)
pwsh .github/skills/release-readiness/scripts/Get-ReleaseReadiness.ps1 \
-SrBranch release/10.0.1xx-sr8 \
-RegressionLabels regressed-in-10.0.70,regressed-in-10.0.80 \
-TrackerKey net10-sr8 \
-OutputDir CustomAgentLogsTmp/release-readiness/sr8
pwsh .github/skills/release-readiness/scripts/Get-ReleaseReadiness.ps1 \
-SrBranch release/10.0.1xx-sr8 \
-Candidate \
-RegressionLabels regressed-in-10.0.80,regressed-in-10.0.90 \
-TrackerKey net10-sr9 \
-OutputDir CustomAgentLogsTmp/release-readiness/sr9-candidate
Preview
pwsh .github/skills/release-readiness/scripts/Get-PreviewReadiness.ps1 \
-Branch release/11.0.1xx-preview6 \
-Mode in-flight \
-TrackerKey net11-preview6 \
-OutputDir CustomAgentLogsTmp/release-readiness/preview6
pwsh .github/skills/release-readiness/scripts/Get-PreviewReadiness.ps1 \
-Branch release/11.0.1xx-preview6 \
-Mode candidate \
-SurveyRef net11.0 \
-TrackerKey net11-preview6 \
-OutputDir CustomAgentLogsTmp/release-readiness/preview6-candidate
The unattended public survey does not know the release-owner-confirmed workload-set
version or private shipping source. It therefore keeps Consumer installability
UNKNOWN rather than guessing that the newest coherent package is the blessed one.
Complete the local gate below before declaring a Preview ready.
Preview: authoritative blessed-build source (.NET Release Tracker)
For Previews, this skill's public survey (CI health + regression classification on net<major>.0 or the preview branch) tells you whether the code is ready, but it cannot on its own name which staged build is the official, blessed preview โ that designation lives in the private .NET Release Tracker plugin. So when answering "run release readiness โฆ is net11 preview6 ready?" / "which build is the official preview6?", consult that authoritative source in addition to running Get-PreviewReadiness.ps1:
-
Classify access first (deterministic gate โ fetches no release data, always exits 0):
pwsh ./.github/skills/dependency-flow/scripts/Get-PreviewReleaseReadiness.ps1
-
Branch on the token:
AVAILABLE_ENABLED โ invoke the dotnet-release-tracker skill for the blessed SDK/runtime + BAR id + stage, and present it as the authoritative official preview build. It is a skill/plugin, not an MCP tool โ don't look for a release-tracker entry in the tool list and give up; run the skill (reload/restart the session if it's enabled but hasn't loaded yet). Combine it with this skill's CI/regression verdict for the full picture.
AVAILABLE_NOT_ENABLED โ the caller has access but the plugin isn't enabled locally; offer the one-time user-scope opt-in, then re-run the gate.
ACCESS_ON_INACTIVE_ACCOUNT โ access exists, but only under a logged-in inactive gh account (named in the gate's inactiveAccount); the plugin loads under the active identity, so advise gh auth switch --user <account> and re-run the gate โ do not invoke the plugin or claim availability under the current identity.
NO_ACCESS โ report from public data only. For the official-build line, fall back to the latest build on the public .NET 11.0.1xx SDK Preview N channel (public BAR/Maestro) and present it labeled as a display-only public-feed candidate โ "source: public preview feed; may not be the final official build." Keep VMR validation UNKNOWN: do not compare/update the MAUI pin or render โ
from that candidate. Don't name or hint at the private tracker tool (see dependency-flow's privacy guardrail).
The full tier table, the user-scope opt-in snippet, and the privacy guardrails live in dependency-flow's "Preview release readiness (authoritative source + access tiers)" section (../dependency-flow/SKILL.md) โ cross-reference it rather than duplicating it here.
Blessed โ green. The release tracker names the official build; it does not substitute for the ship-readiness judgment. A build can be blessed while this skill still reports open regressed-in-* blockers โ surface both.
Don't maintain a standing "๐ท๏ธ Official (blessed) preview build" table in the tracker. The deterministic CI body already owns the public build-pin handling: its "๐ท๏ธ Preview N component build โ branch pins + update paths" section states the pins are explicitly not the blessed build, carries the drift-proof "verify locally" prompt, infers Android/macOS-iOS subscription health from the public PR trail, and identifies VMR as a local official-build reconciliation path. Because the blessed build number is embargoed (withheld from the public issue), a standing public table just renders "๐ withheld" and duplicates that callout. So a local run with tracker access should report the blessed SDK/runtime build in its conversational answer, and only add a line to Release Captain Notes when there's a decision or exception worth persisting โ e.g. the blessed build differs from the branch pin, a promoted build was rejected, or an Android/macOS-iOS subscription is confirmed broken. Don't re-create the section the CI body already renders.
Preview: consumer-installability gate
The branch being green is insufficient: a customer must be able to acquire the
exact SDK workload set, its component manifests, and representative Android,
Apple (including tvOS), Emscripten, MAUI, and runtime packs from a clean source
configuration.
Use the exact workload-set CLI version confirmed by the release owner. Do not
substitute the branch SDK version, and do not assume the newest coherent package
is blessed. Workload-set CLI and NuGet versions have different normalization:
11.0.100-preview.6.26363.2 maps to
11.100.0-preview.6.26363.2 for the NuGet package.
If all assets are public, the confirmed version is enough:
pwsh .github/skills/release-readiness/scripts/Get-PreviewReadiness.ps1 \
-Branch release/11.0.1xx-preview6 \
-Mode in-flight \
-ConfirmedWorkloadSetVersion 11.0.100-preview.6.26363.2 \
'-PublicSafe:$false' \
-OutputDir CustomAgentLogsTmp/release-readiness/preview6-local
If an authenticated shipping feed is required:
-
Create a short-lived PAT at
https://dev.azure.com/dnceng/_usersSettings/tokens.
Select the dnceng organization and grant only Packaging > Read. Use the
shortest practical expiration. Never paste the PAT into a command argument,
NuGet.Config, report, issue, PR, chat transcript, or repository file.
-
Put the credential in NuGet's standard environment variable. The suffix must
exactly match the source name passed to -AdditionalPackageSource.
read -s -p "dnceng Packaging Read PAT: " DNCENG_PACKAGING_PAT; echo
export NuGetPackageSourceCredentials_internal_preview6="Username=release-readiness;Password=${DNCENG_PACKAGING_PAT};ValidAuthenticationTypes=Basic"
unset DNCENG_PACKAGING_PAT
-
Run the local report with the source in name=https://... form:
pwsh .github/skills/release-readiness/scripts/Get-PreviewReadiness.ps1 \
-Branch release/11.0.1xx-preview6 \
-Mode in-flight \
-ConfirmedWorkloadSetVersion 11.0.100-preview.6.26363.2 \
-AdditionalPackageSource 'internal_preview6=<shipping-feed-v3-index-url>' \
'-PublicSafe:$false' \
-OutputDir CustomAgentLogsTmp/release-readiness/preview6-local
-
Use the generated local-only <clear /> NuGet configuration and install
command from preview-readiness.md. Then remove the credential:
unset NuGetPackageSourceCredentials_internal_preview6
-PublicSafe $false intentionally includes exact source URLs and installation
instructions, so keep that output local. The default public-safe report removes
the release-owner-confirmed workload-set version and any candidate version
learned from an authenticated/internal source, including versions repeated in
nested manifest and pack evidence. It also removes additional source names,
URLs, nested source metadata, credentials, and the generated NuGet
configuration. Unconfirmed candidates discovered entirely from public feeds
remain visible as diagnostic evidence.
The gate classifies evidence as follows:
| Installability | Readiness | Meaning |
|---|
installable | READY | Confirmed CLI version, branch pins, required manifests, and representative packs all agree and resolve. |
missing | BLOCKED | A confirmed package or asset is absent from every accessible supplied source. |
mismatched | BLOCKED | The workload set disagrees with the branch SDK, Android, Apple, or runtime pins, or with the target MAUI Preview train. |
unknown | UNKNOWN | Version is unconfirmed, a source is inaccessible, or evidence could not be read. HTTP 401/403 is never treated as proof that a package is missing. |
The isolated source set is deliberate: dotnet-workloads owns the workload-set
package, dotnet<major>-workloads owns platform manifests/assets,
dotnet<major> owns MAUI and Apple manifests/assets,
dotnet<major>-transport owns runtime transport assets, and dotnet-public
plus NuGet.org provide shared dependencies. Do not inherit stale feeds from a
machine-wide NuGet.Config.
Preview: is the branch actually plumbed? (subscription wiring + feed drift)
A preview can pass CI and even have a blessed build yet still not be ship-wired โ
the branch is cut but nothing flows into it, or its promoted feed lags the branch.
The deterministic CI body already gives a best-effort inferred read of the
Android/macOS-iOS wiring from the public PR trail โ the Update path / flow signal
column in its "๐ท๏ธ Preview N component build โ branch pins + update paths" section
(๐ open dep-flow PR / โ
fresh merge โค14d / โ ๏ธ stale >14d / โ none seen). Its VMR
row instead directs the captain to local official-build reconciliation. The
checks below are the authoritative confirmation a local run adds on top of that
inference (darc/BAR can see the subscription itself; CI can only see its PR
exhaust). Run them when the inferred signal is โ ๏ธ/โ, or to confirm a โ
before ship.
A complete "is preview N ready?" answer runs two public BAR/Maestro + git checks
and one access-tiered official-build check alongside the survey:
- Subscriptions wired? Confirm the
release/11.0.1xx-previewN branch has its default-channel mapping and the baseline two subscriptions (android + macios on .NET 11.0.1xx SDK Preview N). There is intentionally no dotnet/VMR subscription: reconcile that pin locally against the official SDK/runtime build because the Maestro channel can differ from the release source of truth. Branch cut + default-channel present but either subscription missing = a start-of-preview flow gap โ surface as an FYI note (not a ship blocker), naming the missing source repos.
- Feed matches the branch? Compare the latest build promoted to the
.NET 11.0.1xx SDK Preview N channel (maestro_latest_build) against origin/release/11.0.1xx-previewN HEAD. Branch ahead of the promoted build = stale feed โ flag it.
- Component pins coherent? Report which
dotnet/android, dotnet/macios, and dotnet/dotnet (VMR) builds MAUI bundles (version + SHA from eng/Version.Details.xml). For Android/macOS-iOS, verify the pin belongs to the component's same-named release/...-previewN branch and carries the expected stage/band; the component branch being ahead of the pin is an FYI because subscriptions may advance it later. Validate the SDK/VMR pin against the official SDK/runtime build from the release source of truth, not netN.0, component branch tip, or the Maestro preview channel. Android's -ci.main.NN scheme is normal for net11 and is not itself an anomaly.
All three checks โ the exact MCP/darc/git commands, the interpretation tables, the
remediation (combined-PR pattern), and live worked examples โ live in dependency-flow's
"Wiring checks: is Preview N actually plumbed?" subsection (Checks A/B/C)
(../dependency-flow/SKILL.md); run them from there and
fold the results into the preview report (missing subs โ an FYI note; stale feed โ a
flagged concern; component pins โ coherent โ
or a flagged divergence) rather than
duplicating the mechanics here. Persist a line in Release Captain Notes only when
the authoritative check diverges from or refines the CI-inferred Flow signal โ e.g.
inferred โ
but the sub points at the wrong channel, or inferred โ ๏ธ/โ confirmed as a
real gap with the missing source repos named. When the local check simply agrees with
the inferred signal, report it conversationally and leave the tracker to the CI body.
Preview action ordering (newly cut branch)
When the preview branch exists but its plumbing is incomplete, present the remediation
as a dependency-ordered sequence. Do not sort unrelated BLOCKED/WATCH rows ahead
of these prerequisites:
- Add the Preview N default-channel mapping.
- Add the baseline Android and macOS/iOS subscriptions. Wait for the
maestro-configuration PR to merge into production, then verify BAR ingestion
with darc get-subscriptions --target-repo https://github.com/dotnet/maui --target-branch <preview-branch>.
- Reconcile MAUI's SDK/VMR pin locally with the official Preview N build and
open the resulting focused component-bump PR. Do not add a VMR subscription.
- Build branch HEAD and promote the resulting MAUI build to the Preview N channel.
- Clear current-preview CI/device/UI failures and finish release validation.
Keep the report scoped to Preview N. Do not mention the netN.0 Preview N+1
bump, main โ netN.0 PRs, the Preview N+1 milestone, or any other next-preview
work. Those belong only in the Preview N+1 candidate/in-flight readiness report.
Parameters
Find-ReleaseReadinessTrackers.ps1
| Parameter | Default | Description |
|---|
-MajorVersion | 0 (auto from eng/Versions.props) | Single major to scan. |
-AllActiveMajors | off | Scan every active major (current + lower in-flight). Mutually exclusive with -MajorVersion. |
-Repo | cwd | Path to a checkout of dotnet/maui. |
-ActivityWindowDays | 7 | Recent-commit window used to compute recentCommitCount. |
-NoFetch | off | Skip git fetch (faster re-runs). |
-OutputJson | โ | File to write the tracker envelope JSON. |
-MaxBranches | 50 | Safety cap on how many SR/preview branches to enumerate per major. |
Get-ReleaseReadiness.ps1 (SR)
| Parameter | Required | Default | Description |
|---|
-SrBranch | Yes | โ | SR branch name (e.g. release/10.0.1xx-sr8). In -Candidate mode, pass the prior SR โ it's the exclude baseline for "what's new". |
-Candidate | No | off | Pre-flight mode โ survey main (with -SrBranch as the prior-SR baseline) to show what WOULD ship in the next SR. |
-Shipped | No | off | Post-ship mode for any explicitly requested tagged SR. Uses immutable stable-tag contents while keeping operational checks live; scheduled trackers normally emit only the most recently tagged SR. |
-ShippedTag | No | latest local stable tag in the SR patch range | Explicit immutable shipped-anchor override for deterministic/manual runs. |
-InheritFromPriorSr | No | off | In -Candidate mode, model the workflow where the prior SR is merged into the new branch after cut. Candidate's "what's shipping" set = main-since-priorSR โช priorSR-only commits. |
-RegressionLabels | One of these | โ | Comma-separated regressed-in-* labels. |
-InferRegressionLabels | One of these | off | Auto-infer from -SrBranch. Agent should confirm before relying on this for automation. |
-Repo | No | dotnet/maui | Repository in owner/name form. |
-MainBranch | No | main | Stable branch used for ancestry checks. |
-ExcludeBranches | No | origin/main | Branches to exclude from SR-only commit computation. |
-Phase | No | all | , , , , or . |
Get-PreviewReadiness.ps1 (Preview)
| Parameter | Required | Default | Description |
|---|
-Branch | Yes | โ | Preview branch name (e.g. release/11.0.1xx-preview6). Required even for candidate runs โ used to derive milestone, tracker key, and regression labels. |
-Mode | No | in-flight | in-flight (survey the preview branch itself) or candidate (survey -SurveyRef instead โ typically net<major>.0). |
-SurveyRef | No | computed | Ref to actually survey. Defaults to $Branch for in-flight; net<major>.0 for candidate. |
-Repository | No | dotnet/maui | Repository in owner/name form. |
-TrackerKey | No | derived | Canonical key (default: net<major>-preview<N>) embedded for idempotent issue lookup. |
-OutputDir | No | โ | If set, writes preview-readiness.{json,md}. |
-OutputFormat | No | markdown | markdown, json, or both. |
-IncludeInternal, -InternalBuildId | No | โ | Release-captain only โ augments report with internal pipeline status when AzDO auth is available. |
-PublicSafe | No | $true | Sanitizes private/internal coordinates from Preview Markdown and JSON. |
-ConfirmedWorkloadSetVersion | No | โ | Exact release-owner-confirmed workload-set CLI version. Required before Consumer installability can become READY. |
-AdditionalPackageSource |
Outputs
| File | Producer | Purpose |
|---|
trackers.json | Find-Trackers | List of active tracker descriptors with detection evidence (one envelope per major) |
release-readiness.{json,md} | Get-ReleaseReadiness | Full SR readiness report |
sr-source-prs.txt | Get-ReleaseReadiness | Flat newline-delimited source PR list; use grep -qxF NNNNN file for instant cherry-pick verification |
sr-commits.json | Get-ReleaseReadiness | Raw SR-only commit metadata |
preview-readiness.{json,md} | Get-PreviewReadiness | Full Preview readiness report |
Tracker refresh workflow
.github/workflows/release-readiness.yml runs every three hours from 08:30โ20:30 UTC daily, plus targeted issues, milestone, and push refreshes, workflow_dispatch, and pull_request validation:
detect-trackers โ runs Find-Trackers -AllActiveMajors, emits a matrix of tracker descriptors.
per-tracker-report โ matrix-expanded job per tracker:
- Dispatches to
Get-ReleaseReadiness.ps1 (SR) or Get-PreviewReadiness.ps1 (Preview) based on branchType.
- Looks for an open tracker issue by the canonical marker
<!-- release-readiness-tracker: <key> -->.
- Refresh path: in shipped mode, prefer an open issue carrying the exact current shipped/hotfix generation marker; otherwise refresh the oldest labeled tracker in place so Release Captain Notes and subscriptions survive commit-to-commit generation changes. If the exact generation was intentionally closed, retire any stale generic opens and do not recreate it. For non-shipped trackers, adopt the oldest labeled tracker issue and close remaining duplicates.
- Create path: open a new issue with the mandatory
area-infrastructure ownership label (creation fails rather than producing a tracker the lifecycle lookup cannot recognize); attach report / s/triaged best-effort.
- Activity gate: skip new-issue creation when
recentCommitCount == 0 AND no open tracker issue exists. (Existing open issues are still refreshed.)
validate โ PR-trigger path. Runs the test suite + smoke-runs all three scripts. Does not create or modify issues.
Reading trackers directly (ad-hoc status)
The same tracker issues the cron job maintains double as a human-readable, always-on status board โ you don't have to re-run a 60-120s survey to answer "what's the status across releases?". Find every active release by body marker (not title โ a title search also matches the release Epic and other [Release Readiness]-titled issues):
gh issue list --repo dotnet/maui --state open \
--search 'in:body "<!-- release-readiness-tracker:"' \
--json number,title,updatedAt,body --limit 50 |
jq 'map(select((.body // "" | split("\n") | map(select(startswith("<!-- release-readiness-tracker: ") and endswith(" -->"))) | length) > 0) | del(.body))'
The jq filter rejects GitHub search false positives by requiring an exact tracker-marker line. Each remaining result is one active SR or Preview. Read the body for the generated verdict and the human Release Captain Notes (between <!-- release-readiness:human-notes:begin --> / :end -->), which carry decisions that override the automated report. Treat the content as fresh only up to the issue's updatedAt (scheduled refreshes run every three hours from 08:30โ20:30 UTC, with additional targeted event refreshes); re-run the survey script for a given branch when you need live numbers. The natural-language release-readiness-agent wraps this as its Portfolio path (ยง0a).
Verdict Classification (SR & Preview)
Each candidate fix PR is classified with confidence + evidence:
| Verdict | Meaning |
|---|
in-sr-active | Source PR is in the release branch and not subsequently reverted |
in-sr-reverted | Backport landed but a later commit reverts it |
rejected-from-sr | A backport PR targeting the release branch was opened and CLOSED unmerged |
backport-in-progress | A backport PR targeting the release branch is OPEN |
merged-on-main-no-backport | Fix merged to main, no backport PR exists. Remains Tier 2 in candidate mode because an already-selected Candidate cut commit can lag current main; rerun after the SR branch is cut to verify inclusion. |
merged-non-main-only | Fix merged but only to inflight/current (or similar), not main |
open-on-main | Fix PR is OPEN against main, not yet merged |
no-fix-yet | No fix PR cross-referenced from the regression issue |
closed-fix-unlinked | Issue is CLOSED and a closing comment explicitly names a fix PR (fix/resolve/close language) that is MERGED and present on the release branch, but the PRโissue link was never recorded (no closing keyword / cross-reference). A bare mention of a PR (e.g. naming the cause PR for context) does not qualify. Non-blocking; action is to add a closing reference for traceability |
needs-human-review | Evidence is contradictory or weak |
SR backport handoff
This skill remains report-only: it MUST NOT post a comment, create a branch, or open a backport PR. When reporting a backport candidate:
-
For merged-on-main-no-backport, include this exact command in the recommendation for the merged source PR:
/backport to release/<major>.0.1xx-sr<N>
-
For open-on-main, wait for the source PR to merge, then recommend the same command. For backport-in-progress, do not trigger a duplicate backport.
-
For merged-non-main-only, do not recommend the command yet. Require a fix PR to merge into main first, then rerun readiness; only recommend the command after the merged source PR's ancestry is verified on main.
-
If the automation reports a conflict, recommend manually cherry-picking the source PR's merge commit onto the SR branch with git cherry-pick -x, resolving and testing the conflict, then opening a PR targeting the SR branch.
Read Gotcha #4 for the generated PR shape and manual fallback commands.
CI Status Categories
| CI verdict | Meaning |
|---|
green | Latest build on the survey ref succeeded across all pipelines |
red-needs-review | Latest build failed or partially succeeded โ investigate failures before judging ship-readiness |
stale | Latest build is older than the survey ref HEAD โ must re-run before judging |
partial-unknown | At least one pipeline couldn't be queried, but no queried pipeline is red or stale |
unknown | No pipeline result could be classified |
Ship-readiness checks (Get-ReleaseReadiness.ps1)
The SR readiness report rolls operational checks into a single Blocking summary at the top, so a release captain sees what must clear before ship without scrolling. Each check emits READY, WATCH, BLOCKED, CLEANUP, or UNKNOWN (CLEANUP = post-release housekeeping that does not block the current ship):
| Check | When | Status meanings |
|---|
Versions.props bump | All SR runs | BLOCKED if eng/Versions.props on main hasn't been bumped past the current SR cycle (next SR has nowhere to flow). |
Versions.props servicing flip | Live-SR mode only | BLOCKED if the SR branch's eng/Versions.props is not flipped to servicing-release mode (PreReleaseVersionLabel=servicing + StabilizePackageVersion=true). After the last backport, create a focused SR PR that preserves PatchVersion, replaces ci.main and its inflight conditional with servicing, and enables stable versions; rerun final CI after it merges. |
Bug template lists SR version | All SR runs | CLEANUP if .github/ISSUE_TEMPLATE/bug-report.yml on main is missing an entry for the SR being shipped (users can't file bugs against the version) โ post-release housekeeping, not a ship blocker. |
Main bumped to next SR cycle | All SR runs | BLOCKED if the next SR cycle's version hasn't been promoted on main. The next action emits the exact one-line PatchVersion edit and PR title while preserving the mainline SDK-band and prerelease settings. |
BAR default-channel mapping | SR branches matching release/X.Y.Zxx-srN | BLOCKED if the SR branch is not wired to the .NET <band> SDK channel in BAR. UNKNOWN if darc isn't on PATH (report includes the exact verification command). |
BAR build for SR HEAD | SR branches with the SR HEAD SHA resolved | READY if BAR has a published build for the SR HEAD commit. WATCH (not blocking โ transient) if CI hasn't published one yet. UNKNOWN if isn't on PATH or the build lookup fails (report includes the exact verification command). |
Next-cycle main bump workflow
After release/<major>.0.1xx-sr<N> is cut, main must move to the next SR
cycle before SR ships. Use a focused PR targeting main:
- Change only
PatchVersion in eng/Versions.props from the current SR
patch to (N+1)*10.
- Title the PR
Update PatchVersion from <old> to <new>.
- Keep
SdkBandVersion, PreReleaseVersionLabel=ci.main, and
StabilizePackageVersion=false unchanged.
- Keep this PR separate from the release branch's servicing-flip PR.
For SR9 โ SR10, the complete source change is:
- <PatchVersion>90</PatchVersion>
+ <PatchVersion>100</PatchVersion>
This is the same one-file, one-line pattern used for SR8 in
#35433 (70 โ 80) and SR9 in
#35879 (80 โ 90). The skill
remains report-only: it must explain this PR precisely, never edit or push
main itself.
Expected ship date
The header line Expected ship date is rendered from Get-ExpectedShipDate, which reads PatchVersion from the survey ref's eng/Versions.props and applies the .NET release cadence:
| PatchVersion | Cadence | Example |
|---|
Multiple of 10 (80, 90, 100โฆ) โ also previews (patch=0) | 2nd Tuesday of the month | SR8 (10.0.80) โ next 2nd Tuesday |
Anything else (81, 82, 91โฆ) | ASAP โ no fixed cadence | SR8 hotfix 10.0.81 โ as soon as ready |
Surfaced in JSON as expectedShipDate.{cadence, date, daysFromNow, formattedLong, note, patchVersion} so downstream automation doesn't redo the math.
Pre-release train cadence (no preview8)
.NET ships a single ordered pre-release train per major:
preview1 โ preview2 โ โฆ โ preview7 โ rc1 โ rc2 โ GA
preview7 is the FINAL preview. There is no preview8. Verified against dotnet/maui's own tags and milestones:
| Major | Last preview | Then | Milestones |
|---|
| .NET 9 | 9.0.0-preview.7.24407.4 | 9.0.0-rc.1.24453.9 โ 9.0.0-rc.2.24503.2 | โ |
| .NET 10 | 10.0.0-preview.7.25406.3 | 10.0.0-rc.1.25424.2 โ 10.0.0-rc.2.25504.7 | .NET 10.0-preview7 โ .NET 10.0-rc1 โ .NET 10.0-rc2 |
Two consequences the report must get right:
- Roll-forward milestone. The cycle after
preview7 is .NET <major>.0-rc1, not .NET <major>.0-preview8. Get-PreviewTrainMilestoneTitle in Get-ReleaseReadiness.ps1 owns this mapping (ordinals 1..7 โ previews, 8 โ rc1, 9 โ rc2, 10+ โ $null); $script:FinalPreviewNumber is the single constant to change if the cadence ever moves. (.NET 5 shipped 8 previews; the 7-preview cadence has held for every major since .NET 6.)
preview7 is the feature/API-lock gate. Because no eighth preview exists, work that misses the preview7 cut does not roll into "the next preview" โ RC is normally API-locked and go-live-licensed, so in practice it slips to the next major. When reporting on a preview7 branch or candidate, say so explicitly: public-API PRs still open at the preview7 cut are a decide-now item, not a defer-later one.
Where the rc mapping actually fires (and where it doesn't). Get-MilestoneHygieneChecks lives only in the SR lane (Get-ReleaseReadiness.ps1); the automated preview lane (Get-PreviewReadiness.ps1 โ Test-PreviewMilestoneExists) validates only the current preview's own milestone and never derives a next-cycle title. So an rc-shaped milestone title is produced only when the SR-lane hygiene check is driven with a preview-shaped branch โ an ad-hoc Get-ReleaseReadiness.ps1 -SrBranch release/<major>.0.1xx-preview7 run, or candidate mode off preview7. An in-flight survey of an actual release/*-rc1 branch produces no milestone checks at all: rc-shaped branches match neither the SR nor the preview shape, so the parser skips them (scenario M15). Treat the previewโrc mapping as defensive normalization for preview-lane inputs, not a live milestone gate on cut rc branches.
Maestro / BAR check gating
The BAR checks shell out to darc (cached probe via Get-Command darc). When darc isn't installed (most CI environments), both checks emit UNKNOWN with the exact local-verification command embedded in the row's Next action โ so the report never silently skips them. The release-readiness agent runs the same checks via the maestro_* MCP tools when the script reports UNKNOWN.
Methodology
Seven critical gotchas this skill encodes โ see references/methodology.md for the full discussion:
-
Cherry-pick number swap: SR backports get NEW PR numbers (e.g. main #35356 โ SR7 #35428). Cannot naively grep source PR numbers; must walk SR-only commits and extract refs from commit bodies.
-
Timeline cross-references: closedByPullRequestsReferences returns empty for most MAUI issues. The skill walks gh api repos/.../issues/N/timeline filtering on cross-referenced events.
-
Forward-flow / non-main merges: A fix can merge into inflight/current only, not main (real example: PR #35609). The skill checks git merge-base --is-ancestor $mergeCommit origin/main before claiming a fix is "on main, just needs backport".
-
Source-to-SR backport workflow: Only a merged fix whose merge commit is on main is ready for the /backport to release/<branch> automation. Non-main fixes must flow to main first; an automated conflict requires a manual git cherry-pick -x of the source merge commit.
-
Servicing flip workflow: The release branch must produce stable packages before ship. All .NET 10 SRs set PreReleaseVersionLabel=servicing and StabilizePackageVersion=true; a focused flip PR is the normal pattern, while SR8 validly inherited those values during its catch-up merge.
-
Next-cycle main bump workflow: After an SR branch is cut, main advances through a separate one-line PatchVersion PR. The report emits the exact old/new XML and title while preserving SdkBandVersion and CI prerelease settings.
-
Default-channel โ per-build feed โ ship Assessment: An SR branch needs a BAR default-channel mapping so its build is promoted and generates the per-build darc-pub-dotnet-maui-<sha8> NuGet feed. The DevDiv ship Assessment must link that feed so CSI/customers can validate the exact candidate packages; without the mapping + promotion the feed never exists and the Assessment ships incomplete (the SR9 miss). The report derives and surfaces the exact feed URL once a promoted build exists.
Shared module
This skill depends on .github/scripts/shared/MauiReleaseVersioning.psm1 for canonical milestone/version parsing (e.g. Get-CurrentMajorVersion, ConvertBranchToMilestone, Get-MilestoneSortKey, Compare-MauiMilestone). The module is also consumed by Fix-MilestoneDrift.ps1 to keep milestone classification consistent across all release-related automation.
scripts/NightlyFeed.ps1 (nightly dogfood feed staleness banner)
Both engines dot-source scripts/NightlyFeed.ps1 to surface a one-line nightly dogfood feed freshness banner at the top of each tracker (just under Generated). The point of the banner is to make it obvious when the dogfood bits people are told to test have stopped flowing โ e.g. when the ci.inflight pipeline is red, the feed goes stale and the banner turns โ so consumers don't waste time validating against builds that never updated.
Key functions (all PURE except the one network call, which is fail-open โ any feed error returns $null / renders a muted "freshness unknown" note and never breaks tracker generation):
| Function | Purpose |
|---|
Get-NightlyFeedFreshness | Queries an Azure Artifacts NuGet feed (dotnet10, dotnet11, โฆ) for the newest published build matching a version-prefix regex; returns version + publish date. Injectable -Fetcher for offline tests. |
Resolve-NightlyDogfoodFreshness | Picks the stream that matters: ci.inflight first (the "shipping next" dogfood bits), the lane band only as a fallback. Conservatively returns matched=$false when only ci.main exists, so a daily main build never paints a false green. |
Format-NightlyFeedLaneLabel | PURE builder for the [`feed`](url) ยท <typeNote> lane label. Centralizes the honest-labeling rule (inflightโci.inflight; bandโcaller-formatted band note; unknownโci.inflight) so the SR and Preview lanes can't drift. |
Get-NightlyFeedTier / Format-NightlyFeedBanner | Bucket age into โ
โค2d ยท โ ๏ธ 3โ6d ยท โ โฅ7d and render the markdown banner. Both take an explicit -Now, so they're deterministic and unit-testable offline. |
Determinism / idempotency: the engine captures one UtcNow per run ($Data['nightlyFeedNow']) and reuses it for both the rendered banner and the semantic-hash tier, so a quiet SR tracker still refreshes when the feed crosses a tier boundary, but a same-tier day-count tick does not churn the issue. The freshness band is folded into Get-ReportSemanticHash (tier|version only โ the raw timestamp is never hashed).
When the banner is โ (feed STALE): the dogfood bits have stopped flowing because the nightly official signed build is failing โ pipeline dotnet-maui (definition 1095, org dnceng / project internal), defined by eng/pipelines/ci-official.yml and scheduled daily on inflight/current. See the azdo-build-investigator skill's Nightly / Official Signed Build (inflight dogfood feed) section for the investigation recipe and the recurring vs-workload.props (MSB4019) failure in the Pack Windows โ "Build Workloads, Sign & Publish" step.
Integration
- Custom agent:
.github/agents/release-readiness-agent.agent.md wraps this skill โ handles regression-label confirmation, runs the script, then uses WorkIQ to add context for rejected-from-sr PRs.
- WorkIQ: NOT called from the PowerShell scripts (PowerShell can't invoke MCP tools). The agent enriches the script's JSON output with WorkIQ context where needed.
- .NET Release Tracker (Preview blessed build): for Previews, the authoritative official build is designated in the private
dotnet-release-tracker plugin, reached through dependency-flow's deterministic access gate (Get-PreviewReleaseReadiness.ps1). See Preview: authoritative blessed-build source above; the tier table + opt-in live in ../dependency-flow/SKILL.md.
Anti-Patterns
โ Don't naively grep source PR numbers in the SR git log. The backport PR number replaces the source PR number in the merge commit subject. Use sr-source-prs.txt (produced by this skill) instead.
โ Don't claim a fix is on main based on pr-view --state MERGED. PRs can be merged into inflight/current only. The skill's onMain field is the authoritative check.
โ Don't trust issue-title similarity. Two issues can have nearly identical titles and refer to different platform-specific regressions (e.g. #35313 is the Android version, #35326 is the iOS/Mac/Win version with a different fix path). Always filter by the regressed-in-* label, not by title.
โ Don't run with -InferRegressionLabels for automated workflows without surfacing the inferred labels for confirmation. Label inference is brittle for non-standard SR cycles.
โ Don't infer "in-flight" from branch existence alone. The detector uses the tag-existence rule โ a release is in-flight if and only if its expected tag has not been published. Branches can linger after their release ships (and SR branches don't exist yet for SR candidates).
Tests
pwsh -NoProfile -File .github/skills/release-readiness/tests/Test-ReleaseReadiness.ps1
The harness covers:
- Lane 1โ4 detection (shipped patch set, SR-from-main candidate, in-flight SR branches, preview lane) against the live
dotnet/maui clone
- Tracker emission for SR2/SR3 (inactive), SR8 (active in-flight), SR9 (active candidate), and net11 preview6 (active candidate)
-AllActiveMajors end-to-end across net10 + net11 with the expected tracker counts
Get-ReleaseReadiness verdict classification using known-answer data from the SR7 readiness analysis (e.g. #35313 โ in-sr-active, #35344 โ in-sr-active via the SafeArea follow-on fix, #35771 โ no-fix-yet)
- Idempotent body hash stability across re-runs โ SR trackers only (the scheduled/event-driven workflow compares the embedded
<!-- release-readiness-hash: sha=... --> marker against the live issue and skips the edit when the semantic content is unchanged, so re-runs don't churn the tracker). Preview trackers carry no hash marker and are refreshed on every scheduled run.
- Nightly dogfood feed banner (
NightlyFeed.ps1) โ offline unit coverage for the lane-label honest-labeling rule (Format-NightlyFeedLaneLabel), the ci.inflight-first / ci.main-false-green resolver, ageโtier bucketing, the fail-open feed query (mocked -Fetcher), and the banner's fold into Get-ReportSemanticHash (tier change refreshes, same-tier day tick does not). All network-free via injected fixtures and explicit -Now.
- Preview consumer installability (
PreviewInstallability.ps1) โ offline fixtures cover CLI/NuGet version conversion, workload-set discovery with MSI exclusion, branch-pin coherence, source-role resolution, real manifest alias-to resolution for Android/Emscripten/runtime representative packs, platform prerequisites, isolated <clear /> configuration, malformed/401/403=UNKNOWN semantics, verdict mapping, and public-output redaction.