<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#39962
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [x] Timeouts are implemented and retries are limited to avoid infinite
loops
- [ ] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes
## Testing
- [x] Added/updated automated tests
- [ ] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [x] QA'd all new/changed functionality manually
## Database migrations
- [x] Checked schema for all modified table for columns that will
auto-update timestamps during migration.
- [ ] Confirmed that updating the timestamps is acceptable, and will not
cause unwanted side effects.
- N/A
- [x] Ensured the correct collation is explicitly set for character
columns (`COLLATE utf8mb4_unicode_ci`).
## New Fleet configuration settings
- [ ] Setting(s) is/are explicitly excluded from GitOps
If you didn't check the box above, follow this checklist for
GitOps-enabled settings:
- [x] Verified that the setting is exported via `fleetctl
generate-gitops`
- [x] Verified the setting is documented in a separate PR to [the GitOps
documentation](https://github.com/fleetdm/fleet/blob/main/docs/Configuration/yaml-files.md#L485)
- [x] Verified that the setting is cleared on the server if it is not
supplied in a YAML file (or that it is documented as being optional)
- [x] Verified that any relevant UI is disabled when GitOps mode is
enabled
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added “Patch when closed” deployment policies to update software only
when the application is not running.
* Added deployment controls for force install, patching, and manual,
forced, or closed-app patch options.
* Fleet-maintained apps now automatically detect whether the application
is open.
* GitOps configurations support patch-when-closed settings with
validation.
* **UI Improvements**
* Added clear activity and installation messages when updates are
skipped because an app is open.
* Replaced the Patch action with a unified Deploy workflow.
* **Bug Fixes**
* Prevented skipped updates from being incorrectly retried as failed
installations.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Issue
Resolves#50528.
## Description
On the My device > Self-service page, with a category selected and a
search query typed, the "Install all" button previously ignored the
search: it counted (and queued) every uninstalled item in the category,
including software the search had filtered out.
This PR scopes the button — count *and* install target — to the visible
subset:
- **Backend:** `POST /device/{token}/software/install_all` now accepts a
`query` param. It's threaded through
`SelfServiceInstallAllSoftwareTitles` → `GetSoftwareTitlesForInstallAll`
→ `opts.ListOptions.MatchQuery` on `ListHostSoftware`, reusing the same
LIKE-on-`software_titles.name` semantics as the self-service list
endpoint.
- **Frontend:** new `filterSoftwareByQuery` helper layers on top of the
category filter to drive `uninstalledCount` / `hasInProgress` and the
value sent to install_all. Empty queries are stripped so the API isn't
called with `?query=`.
`display_name` matching is deliberately out of scope — the search filter
across BE list, desktop table, and mobile filter is all raw-`name`-only
today, so broadening install_all alone would re-introduce a similar
mismatch. Filed as a follow-up: #50750.
## Screen recording
In recording:
- (FE fix) showing that the UI is filtering out install all count to be
only what's on the screen
- (BE fix) showing that the call to the API only queues up the install
all for the installers shown on the screen when clicked
https://github.com/user-attachments/assets/aaae3d29-dccf-484d-910f-67ca335bf0e8
## Testing
- FE unit tests: `filterSoftwareByQuery` helper, `SelfServiceCard`
count-with-query + POST-with-query, `InstallAllInCategoryButton` prop
forwarding.
- BE unit test: EE service forwards the match query to the datastore.
- BE datastore test: query, category+query, empty-match cases.
- BE integration test: new "scopes to the query parameter when provided"
subtest in `TestInstallAllSelfServiceSoftware`.
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* “Install all” now respects the active self-service search query.
* Counts, progress indicators, and installation requests now reflect
only software matching the current search and category filters.
* Empty or whitespace-only searches continue to include all software in
the selected category.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Software title details now return installer scripts and managed app
configuration only to users who can read the installer. Uninstalling
software from the My device page now applies the same self-service and
label scope rules as installing.
# Checklist for submitter
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Security & Permissions**
* Restricted installer scripts, managed-app settings, and related
configuration to authorized viewers.
* Preserved package metadata while hiding sensitive installation details
from unauthorized roles.
* Improved access handling for requests without an assigned team or
involving inaccessible fleets.
* **Bug Fixes**
* Updated device-initiated software removal to honor self-service
eligibility and label scope.
* Added clearer errors when software is unavailable for self-service or
outside the device’s scope.
* Prevented software titles from inaccessible fleets from appearing in
results.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#45728
Changes:
- Adds a new redis key to keep track of downloaded packages. It starts
out with an empty list and gets filled with each download. Each update
writes the entire struct at once to the key.
- Adds logging in the fleetctl gitops client to show which packages were
downloaded
- Fixes the categories key potentially expiring
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [ ] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- ❌ Timeouts are implemented and retries are limited to avoid infinite
loops
- Right now the batch will write the whole slice of all packages to a
single redis key for every package in the loop. Looks like performance
is acceptable for now (500 packages), but maybe this will need to be
limited.
- [ ] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes
## Testing
- [x] Added/updated automated tests
- [ ] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [x] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
## New Features
- Added per-package software download progress in fleetctl GitOps.
- Progress now reports downloading, completed, skipped, and failed
packages during real and dry runs.
- Installation output now distinguishes applying and applied stages.
## Bug Fixes
- Improved download error messages and cached-package handling.
- Prevented duplicate progress messages and ensured tracking issues do
not interrupt successful software batches.
## Tests
- Expanded coverage for progress reporting, failures, dry runs, package
types, and authorization scenarios.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
**Related issues:** Resolves#50068, Resolves#50106, Resolves#50107,
Resolves#50108, Resolves#50110, Resolves#50114
Follow-on fixes from QA of #41470 (Python script-only packages):
- Software-installer validation errors are action-neutral, so the Add
and Edit flows each show the correct single verb, and the
unsupported-file error names a content/format mismatch instead of
blaming the extension (#50068, #50107).
- `.py` packages accept `setup_experience_platform` (`darwin`/`linux`),
matching `.sh` (#50106).
- A failed-to-run install script (exit code `-1`) now renders a
diagnostic instead of empty output, and orbit surfaces the underlying
execve error (#50108).
- The install-rejection message for `.sh`/`.py` packages says "macOS and
Linux hosts" instead of "linux" (#50110).
- Orbit writes each script's temp file with an extension matching its
shebang (`.py`/`.sh`/`.ps1`), so tracebacks reference the right file
type (#50114).
# Checklist for submitter
- [x] Changes file added for user-visible changes in `changes/`.
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
## fleetd/orbit/Fleet Desktop
- [x] If the change applies to only one platform, confirmed that
`runtime.GOOS` is used as needed to isolate changes.
- [x] Verified compatibility with the latest released version of Fleet
(orbit-only change; the server↔agent `SoftwareInstallDetails` contract
is unchanged).
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Improved installer validation and rejection messaging for
unsupported/invalid package contents (including correcting “add” vs
“edit” wording and avoiding duplicated phrasing).
* Added clearer diagnostics when install scripts fail to start
(including empty output cases).
* Corrected handling of script-only packages so Python scripts use the
proper script type/extension, reducing misleading tracebacks.
* Updated platform availability messaging so `.sh`/`.py` packages
display macOS+Linux support.
* **New Features**
* Python script-only packages can now specify macOS and Linux setup
experience platforms.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#49416
- Adds patch_when_closed to the new policy, update policy, and update
package endpoints
- Sets pre_install_query in the software title endpoint to
software_installers.app_open_query if the policy is a patch when closed
policy (discussed in standup)
- Deletes the existing pre_install_query if setting patch_when_closed
(discussed in standup)
- Fixes a small error in the patch_policy package
- Fixes the default macOS query in patch_policy package to use to escape
symbols in a.path
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [ ] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [ ] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [ ] Timeouts are implemented and retries are limited to avoid infinite
loops
- [ ] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes
## Testing
- [x] Added/updated automated tests
- [ ] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [x] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added a “patch when closed” option for patch policies and
Fleet-maintained apps.
* Added installer controls for enabling patching and configuring whether
apps must be closed before installation.
* Automatically manages pre-install behavior and continuous automation
requirements for these policies.
* **Bug Fixes**
* Improved app detection based on application path prefixes on macOS.
* Corrected RStudio process detection on Windows.
* Added validation to prevent incompatible patch policy settings.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Resolves#44406
Windows programs report a version in their name (e.g. `Granola
7.373.2`), so each version created its own `software_title` and never
linked to the Fleet-maintained app installer's title (`Granola`), hiding
the uninstall action. macOS handles this via `bundle_identifier`;
Windows had no join key.
- Give matching Windows programs the canonical FMA name at ingestion
(name-prefix match), so all versions collapse onto the title the
installer owns. `software.name` is unchanged.
- Merge already-mismatched versioned titles onto the canonical title in
`ReconcileMaintainedAppSoftwareNames` (runs on FMA sync; no migration
needed).
---------
Co-authored-by: Tim Lee <timlee@fleetdm.com>
Co-authored-by: Juan Fernandez <juan@fleetdm.com>
**Related issue:** Resolves#49418
Adds `patch_when_closed` support to GitOps for patch policies and
round-trips it through `fleetctl generate-gitops`.
- Validate `patch_when_closed` in the patch-policy YAML: reject an
explicit `continuous_automations_enabled: false` alongside it (GitOps is
declarative — the datastore would otherwise silently force it on),
auto-set it when omitted, and reject a `pre_install_query` on the
referenced Fleet-maintained app (Fleet manages that query).
- Emit `patch_when_closed` from `fleetctl generate-gitops`.
# Checklist for submitter
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
## New Fleet configuration settings
Follow this checklist for GitOps-enabled settings:
- [x] Verified that the setting is exported via `fleetctl
generate-gitops`
- [x] Verified the setting is documented in a separate PR to [the GitOps
documentation](https://github.com/fleetdm/fleet/blob/main/docs/Configuration/yaml-files.md)
(#49022)
- [x] Verified that the setting is cleared on the server if it is not
supplied in a YAML file
---------
Co-authored-by: Jonathan Katz <yehonatankatz@gmail.com>
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#49511 and #46837 as a whole
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [x] Timeouts are implemented and retries are limited to avoid infinite
loops
- [x] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes
## Testing
- [x] Added/updated automated tests
- [x] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [x] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added support for Fleet built-in variables in host scripts, software
installer scripts, setup-experience scripts, and maintained-app
installer scripts.
* Variables are resolved per host at execution time; saved content
remains unexpanded.
* **Bug Fixes**
* Requests now validate Fleet variables up-front, with clear
script-specific error messages for unsupported variables.
* Added improved messaging when variable resolution fails during
execution.
* Enforced Fleet Premium licensing for script/installer flows that use
Fleet variables.
* **Documentation**
* Documented supported variables and Premium requirements, including
usage examples.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
**Related issue:** Resolves#49455
Offer `.py` script-only packages on macOS hosts, matching `.sh`.
# Checklist for submitter
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Python (`.py`) installer packages are now treated as compatible
alongside shell (`.sh`) installers on macOS and Linux.
* Python installers can now appear in software availability,
self-service installation, and setup experience selections.
* Windows behavior remains unchanged (Unix-script installers are
excluded).
* **Bug Fixes**
* Improved cross-platform compatibility matching for Unix-like hosts
when choosing the first eligible installer package.
* **Tests**
* Added and expanded unit/integration coverage for `.py` installer
compatibility across platforms and flows.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
**Related issue:** Resolves#41470
Adds support for uploading Python (`.py`) script-only software packages
— accepted as script-only (the file contents become the install script;
advanced options and automatic install follow `.sh`/`.ps1`), assigned
the new `py_packages` source, and installable on macOS and Linux hosts
across the UI, REST API, and GitOps.
Feature branch combining the backend (#48942) and frontend (#48946)
sub-PRs.
# Checklist for submitter
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added support for Python (`.py`) script-only software packages across
UI uploads, API/self-service installs, and GitOps parsing.
* Python installers now derive metadata correctly and render the proper
Python icon, with install eligibility for macOS & Linux.
* **Bug Fixes**
* Improved installer-script validation and “supported file types” error
messages to include `.py` (and consistent handling of related script
fields/options).
* **Tests**
* Expanded unit, integration, and GitOps tests to cover Python package
parsing, metadata derivation, platform/host eligibility, and UI
rendering.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#44954
# Checklist for submitter
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added Custom Host Vitals management (create/edit/delete, search) with
copyable variable tokens and per-host values.
* Show Custom Host Vitals on host details, including role-based editing.
* Enabled Custom Host Vitals in host-vitals labels and the activity
feed.
* Extended GitOps to manage global Custom Host Vitals declaratively
(including dry-run behavior).
* Split Controls → Variables into Global Variables and Custom Host
Vitals, including routing updates.
* **Bug Fixes**
* Improved validation and expansion of Custom Host Vital references
across scripts, profiles, installers, and deployments, with clearer
failures when values are missing or invalid.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
**Related issue:** Resolves#49234
Editing a software installer to a new version failed with "The selected
package is for different software" when the title's stored
(osquery-reported) name differed from the installer's extracted name.
The edit now validates by software identity (bundle identifier / upgrade
code / name) instead of an exact name match.
# Checklist for submitter
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** #43667
# Summary
Renames the unreleased GitOps field `setup_experience_platforms` to
singular `setup_experience_platform`, accepting a comma-separated string
of `darwin`/`linux` (rejecting the `macos` alias) to match the
query/policy/label `platform` convention.
# Checklist for submitter
- [ ] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
## Testing
- [ ] Added/updated automated tests
- [ ] QA'd all new/changed functionality manually
## New Fleet configuration settings
- [ ] Verified that the setting is exported via `fleetctl
generate-gitops`
- [ ] Verified the setting is documented in a separate PR to [the GitOps
documentation](https://github.com/fleetdm/fleet/blob/main/docs/Configuration/yaml-files.md#L485)
- [ ] Verified that the setting is cleared on the server if it is not
supplied in a YAML file (or that it is documented as being optional)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Improvements**
* Updated software setup-experience platform configuration to use a
single comma-separated `setup_experience_platform` value.
* Platform values are normalized for casing and whitespace,
deduplicated, and validated against supported platforms.
* macOS setup selections now use the canonical `darwin` value; the
`macos` alias is rejected.
* GitOps-generated configurations now use the updated field name and
platform format.
* **Bug Fixes**
* Improved validation messages for invalid setup-experience platform
values.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
**Related issue:** Resolves#43667
# Summary
Adds a `setup_experience_platforms` field to the GitOps software package
spec so `.sh` script-only installers can be selected for macOS setup
experience declaratively. Reconciles the cross-platform selection table
on every batch apply.
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`. See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
## New Fleet configuration settings
- [x] Verified that the setting is exported via `fleetctl
generate-gitops`
- [x] Verified the setting is documented in a separate PR to [the GitOps
documentation](https://github.com/fleetdm/fleet/blob/main/docs/Configuration/yaml-files.md#L485)
- [x] Verified that the setting is cleared on the server if it is not
supplied in a YAML file (or that it is documented as being optional)
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added declarative `setup_experience_platforms` to software package
definitions to control “setup experience” targets, including selecting
script-only installers for macOS (mapped appropriately).
* Batch uploads now propagate these cross-platform selections and
reconcile installer cross-entries.
* **Bug Fixes**
* Improved platform normalization (trimming, casing, alias mapping),
deduplication, and extension-specific validation.
* Enhanced update behavior: omitting the field leaves existing
selections unchanged; providing an empty list clears them, with correct
setup/installation timing.
* **Tests**
* Added unit and integration coverage for normalization and batch
re-apply/reconcile behavior (nil vs empty, idempotency, mixed updates,
and validation failures).
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
**Related issue:** Resolves#48398
Applies first-added-wins install precedence everywhere a host can match
more than one package of a software title: manual install, self-service,
policy auto-install, setup experience, and the host-software read all
resolve the first-added package the host is in label scope for. Policy
automations can now target a specific package (defaulting to
first-added), and deleting a package re-points its automations to the
first-added surviving package.
# Checklist for submitter
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements).
## Testing
- [x] Added/updated automated tests
- [x] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [x] QA'd all new/changed functionality manually
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#47679
Adds a few things:
- exports pinned version in `generate-gitops` enclosed in double quotes
- adds `pinned_version` to the edited software activity. When set to a
full or major version it shows up in details, when set to latest or
unchanged it shows up as `pinned_version: null` (some other fields like
display_name also dont show up when unchanged)
- fixes a bug where some FMA's like google chrome couldn't be pinned to
major version because they couldn't be converted to semver (by just
splitting the version on periods instead of converting to semver)
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [ ] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [ ] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [ ] Timeouts are implemented and retries are limited to avoid infinite
loops
- [ ] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes
## Testing
- [x] Added/updated automated tests
- [ ] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [x] QA'd all new/changed functionality manually
## New Fleet configuration settings
- [ ] Setting(s) is/are explicitly excluded from GitOps
If you didn't check the box above, follow this checklist for
GitOps-enabled settings:
- [x] Verified that the setting is exported via `fleetctl
generate-gitops`
- [ ] Verified the setting is documented in a separate PR to [the GitOps
documentation](https://github.com/fleetdm/fleet/blob/main/docs/Configuration/yaml-files.md#L485)
- [ ] Verified that the setting is cleared on the server if it is not
supplied in a YAML file (or that it is documented as being optional)
- [ ] Verified that any relevant UI is disabled when GitOps mode is
enabled
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Software titles now support version pinning in GitOps exports for
fleet-managed applications, with pinned version values properly
formatted as quoted strings in the exported YAML configuration
* Activity logs now record when pinned version information is modified
during software editing operations
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#46726
Updates the `PATCH /api/v1/fleet/software/titles/:id/package` endpoint
to be able to set the pinned version in the `software_title_team_pins`
table and set the active installer if it changed.
`GET /software/titles/:id` returns the pinned version for a title from
that table now.
A few small fixes and updates that are related to this feature but not
strictly in the scope of the subtask are also included (see individual
comments).
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [ ] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [ ] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [ ] Timeouts are implemented and retries are limited to avoid infinite
loops
- [ ] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes
## Testing
- [x] Added/updated automated tests
- [ ] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [x] QA'd all new/changed functionality manually
## New Fleet configuration settings
This setting is already handled by gitops, this PR adds control of it
with the update software installer api
- [ ] Setting(s) is/are explicitly excluded from GitOps
If you didn't check the box above, follow this checklist for
GitOps-enabled settings:
- [ ] Verified that the setting is exported via `fleetctl
generate-gitops`
- TODO in another PR
- [ ] Verified the setting is documented in a separate PR to [the GitOps
documentation](https://github.com/fleetdm/fleet/blob/main/docs/Configuration/yaml-files.md#L485)
- [ ] Verified that the setting is cleared on the server if it is not
supplied in a YAML file (or that it is documented as being optional)
- [ ] Verified that any relevant UI is disabled when GitOps mode is
enabled
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added version pinning for Fleet-maintained app installers, allowing
teams to lock to specific versions or clear pins to use "Latest"
* Policies automatically repoint to the newly active installer when a
version pin is changed
* Support for semantic version syntax including caret-style version
constraints
* **Tests**
* Added comprehensive integration test for Fleet-maintained app version
pinning workflows and validation rules
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#44710
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [ ] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [ ] Timeouts are implemented and retries are limited to avoid infinite
loops
- [ ] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes
## Testing
- [x] Added/updated automated tests
- [ ] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [x] QA'd all new/changed functionality manually
## New Fleet configuration settings
- [ ] Setting(s) is/are explicitly excluded from GitOps
If you didn't check the box above, follow this checklist for
GitOps-enabled settings:
- [ ] Verified that the setting is exported via `fleetctl
generate-gitops`
- [ ] Verified the setting is documented in a separate PR to [the GitOps
documentation](https://github.com/fleetdm/fleet/blob/main/docs/Configuration/yaml-files.md#L485)
- [ ] Verified that the setting is cleared on the server if it is not
supplied in a YAML file (or that it is documented as being optional)
- [ ] Verified that any relevant UI is disabled when GitOps mode is
enabled
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Fixed validation failures when script packages contain spaces in their
names during GitOps operations.
* Script-based installer packages are no longer incorrectly validated as
standard URLs.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves #
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [ ] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- Not needed
- [ ] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [ ] Timeouts are implemented and retries are limited to avoid infinite
loops
- [ ] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes
## Testing
- [x] Added/updated automated tests
- [ ] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [x] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Batch software installer and app-association endpoints now return the
list of referenced self-service categories.
* Category fields support an “omit when unset” JSON behavior so omitted
vs empty categories are distinguishable.
* **Bug Fixes**
* Improved category validation (trim + case-insensitive dedupe) and
GitOps reconciliation to remove unused categories.
* **Chores**
* GitOps schema simplified: no separate top-level
self_service_categories; categories are defined inline with packages.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
# Checklist for submitter
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [x] Timeouts are implemented and retries are limited to avoid infinite
loops
- [x] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
## Database migrations
- [x] Checked schema for all modified table for columns that will
auto-update timestamps during migration.
- [x] Confirmed that updating the timestamps is acceptable, and will not
cause unwanted side effects.
- [x] Ensured the correct collation is explicitly set for character
columns (`COLLATE utf8mb4_unicode_ci`).
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
## Release Notes
* **New Features**
* In-house iOS app manifest and package downloads now use secure
per-install tokens embedded in the URL path instead of query parameters
* Installation tokens are bound to specific devices and teams, enhancing
security
* Installation tokens automatically expire after 6 hours
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Jonathan Katz <yehonatankatz@gmail.com>
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#46393
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [ ] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [ ] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [ ] Timeouts are implemented and retries are limited to avoid infinite
loops
- [ ] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes
## Testing
- [x] Added/updated automated tests
- [ ] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [x] QA'd all new/changed functionality manually
This pull request improves how the platform is determined for software
installers, ensuring that the installer's stored platform is used when
available, rather than inferring it from the file extension. This is
especially important for `.zip` installers, which may be used on
multiple platforms. The changes also add a new test to verify this
behavior.
Trying to install a zip-based FMA on Windows:
<img width="1485" height="499" alt="Screenshot 2026-06-04 at 12 39
50 PM"
src="https://github.com/user-attachments/assets/cbeecefd-6376-44d9-8e93-dd93f0c714f9"
/>
**Platform detection improvements:**
* Added the `installerRequiredPlatform` helper function, which returns
the file extension and uses the installer's stored `Platform` field if
set; otherwise, it falls back to inferring the platform from the
extension. This ensures correct handling for installers like `.zip`
files used on both Windows and macOS.
* Updated all relevant service methods
(`installSoftwareTitleUsingInstaller`, `UninstallSoftwareTitle`, and
`SelfServiceInstallSoftwareTitle`) to use `installerRequiredPlatform`
for platform validation instead of inferring from the file extension
alone.
[[1]](diffhunk://#diff-b3883848dac3454f4ed4968a94f9ca335241f9f99bfad92f95a6bfb7157523ebL1592-R1592)
[[2]](diffhunk://#diff-b3883848dac3454f4ed4968a94f9ca335241f9f99bfad92f95a6bfb7157523ebL1702-R1701)
[[3]](diffhunk://#diff-b3883848dac3454f4ed4968a94f9ca335241f9f99bfad92f95a6bfb7157523ebL3226-R3224)
**Testing improvements:**
* Added a new test, `TestInstallZipInstallerUsesStoredPlatform`, to
verify that `.zip` installers use the stored platform (e.g., "windows")
rather than inferring "darwin" from the extension.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Platform enforcement now prefers stored installer platform metadata
(zip treated as ambiguous), preventing incorrect platform selection for
install, uninstall, and self-service install flows.
* **Tests**
* Added tests verifying `.zip` installers use stored platform metadata
for install, uninstall, and self-service install scenarios.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Carlo <1778532+cdcme@users.noreply.github.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Carlo DiCelico <carlo@fleetdm.com>
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#43729
# Details
Adds output to GitOps runs indicating which custom/FMA software packages
would be deleted. This involves adding a `deleted_packages` key to the
`/software/batch/:request_uuid` ("Get status of software batch-apply
request") API, which will be documented separately.
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [X] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [X] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
## Testing
- [X] Added/updated automated tests
- [X] QA'd all new/changed functionality manually
- [X] verified that a GitOps dry run produces one "would've deleted"
line per custom package / fma that would be deleted
- [X] verified that a GitOps real run produces one "deleted" line per
custom package / fma that was deleted
- [X] verified that adding software is unaffected
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* GitOps batch software operations now report packages pending deletion:
dry-runs show "would've deleted" warnings and real runs show deletions;
apply flows surface per-package deletion messages.
* Empty payload dry-run now still reports pending deletions when
applicable.
* **Tests**
* Added integration and datastore tests validating deletion-warning
output, pending-deletion detection, and related result handling.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
- **Adding vpp users table**
- **Adding ManagedAppleID to HostMDM tables to pull when installing vpp
apps to a BYOD user enrolled device**
- **Adding user create vpp apis**
- **Updating install application to support user enrolled devices**
- **Handling already installed gracefully**
- **Adding provision user logic**
- **Adding logic to associate assets on install**
- **Add license managemnt logic**
- **Adding self service ui**
- **Adding setup experience support**
- **Adding setup experience flow**
- **Fix issue with fleet's docker image in k8s environments (#44373)**
- **Changing how the appleid is captured to use idp mdm users**
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Account-based User Enrollment (iOS/iPadOS) supports VPP and in‑house
.ipa installs, including self‑service installs.
* Enrollment now applies the selected Setup experience automatically for
user‑enrolled hosts.
* Installs and license associations are scoped to account user
associations (user‑scoped installs).
* **Bug Fixes**
* Improved error messaging for license/association failures (including
per‑user device cap cases).
* Mobile UI: self‑service now shows user‑enrolled iOS/iPadOS apps.
* “Already installed” install results are treated as acknowledged
successes.
<!-- review_stack_entry_start -->
[](https://app.coderabbit.ai/change-stack/fleetdm/fleet/pull/45202)
<!-- review_stack_entry_end -->
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
**Related issue:**
Ref #34797
Ref #42675
## Problem
When a software installer spec has no `hash_sha256`, Fleet re-downloads
the package, re-extracts metadata, and re-upserts the DB on every GitOps
run, even if the upstream file hasn't changed. For deployments with 50+
URL-only packages across multiple teams, this wastes bandwidth and
processing time on every run.
## Solution
By default, use etags to avoid unnecessary downloads:
1. First run: Fleet downloads the package normally and stores the
server's ETag header
2. Subsequent runs: Fleet sends a conditional GET with `If-None-Match`.
If the server returns 304 Not Modified, Fleet skips the download,
metadata extraction, S3 upload, and DB upsert entirely
Opt-out with `always_download:true`, meaning packages continue to be
downloaded and re-processed on every run, same as today. No UI changes
needed.
```yaml
url: https://nvidia.gpcloudservice.com/global-protect/getmsi.esp?version=64&platform=windows
always_download: true
install_script:
path: install.ps1
```
### Why conditional GET instead of HEAD
Fleet team [analysis of 276 maintained
apps](https://github.com/fleetdm/fleet/pull/42216#issuecomment-4105430061)
showed 7 apps where HEAD requests fail (405, 403, timeout) but GET works
for all. Conditional GET eliminates that failure class: if the server
doesn't support conditional requests, it returns 200 with the full body,
same as today.
### Why opt-in
5 of 276 apps (1.8%) have stale ETags (content changes but ETag stays
the same), caused by CDN caching artifacts (CloudFront, Cloudflare,
nginx inode-based ETags). The `cache` key lets users opt in per package
for URLs where they've verified ETag behavior is correct.
Validation rejects `always_download: true` when hash_sha256` is set
## Changes
- New YAML field: `cache` (bool, package-level)
- New migration: `http_etag` VARCHAR(512) column (explicit
`utf8mb4_unicode_ci` collation) + composite index `(global_or_team_id,
url(255))` on `software_installers`
- New datastore method: `GetInstallerByTeamAndURL`
- `downloadURLFn` accepts optional `If-None-Match` header, returns 304
as `(resp, nil, nil)` with `http.NoBody`
- ETag validated per RFC 7232 (ASCII printable only, no control chars,
max 512 bytes) at both write and read time
- Cache skipped for `.ipa` packages (multi-platform extraInstallers)
- TempFileReader and HTTP response leak prevention on download retry
- Docs updated in `yaml-files.md`
## What doesn't change
- Packages with `hash_sha256`: existing hash-based skip, untouched
- FMA packages: FMA version cache, untouched
- Packages with `always_download: true`: identical to current behavior
- Fleet UI: no changes
## Test plan
Automated testing:
- [x] 16 unit tests for `validETag`
- [x] 8 unit tests for conditional GET behavior (304, 200, 403, 500,
weak ETag, S3 multipart, no ETag)
- [x] MySQL integration test for `GetInstallerByTeamAndURL`
- [x] All 23 existing `TestSoftwareInstallers` datastore tests pass
- [x] All existing service tests pass
Manual testing:
- [x] E2E: 86 packages across 6 CDN patterns, second apply shows 51
conditional hits (304)
- [x] @sgress454 used a local fileserver tool to test w/ a new instance
and dummy packages
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* ETag-based conditional downloads to skip unchanged remote installer
files.
* New always_download flag to force full re-downloads.
* **Tests**
* Added integration and unit tests covering conditional GETs, ETag
validation, retries, edge cases, and payload behavior.
* **Chores**
* Persist HTTP ETag and related metadata; DB migration and index to
speed installer lookups.
* Added installer lookup by team+URL to support conditional download
flow.
* **Bug Fix**
* Rejects using always_download together with an explicit SHA256 in
uploads.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Scott Gress <scott@fleetdm.com>
Co-authored-by: Scott Gress <scott@pigandcow.com>
Co-authored-by: Ian Littman <iansltx@gmail.com>
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#38988
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [ ] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [ ] Timeouts are implemented and retries are limited to avoid infinite
loops
- [ ] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes
## Testing
- [x] Added/updated automated tests
- [ ] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [ ] QA'd all new/changed functionality manually
**Related issue:** Resolves#42480
Script packages (`.sh`, `.ps1`) are persisted server-side in the
`script_contents` table, but `addScriptPackageMetadata` validates them
with `ValidateHostScriptContents(scriptContents, false)`, enforcing the
unsaved script limit (10K characters) instead of the saved limit (500K
characters).
This one-line fix changes `false` to `true`, allowing script packages up
to 500K characters, consistent with other saved scripts.
# Checklist for submitter
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
## Testing
- [x] Added/updated automated tests
- [ ] QA'd all new/changed functionality manually
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#41385
# Details
This PR updates `fleetctl` to use the new API urls and params when
communicating with Fleet server. This avoids deprecation warnings
showing up on the server that users won't be able to fix.
Most of the changes are straightforward `team_id` -> `fleet_id`. A
couple of code changes have been pointed out. The most interesting is in
icon URLs, which can be persisted in the database (so we'll need to do a
migration in Fleet 5 if we want to drop support for `team_id`.
Similarly the FMA download urls are briefly persisted in the db for the
purpose of sending MDM commands. If we drop team_id support in Fleet 5
there could be a brief window where there are unprocessed commands in
the db still with `team_id` in them, so we'll probably want to migrate
those as well.
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [ ] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
n/a - all internal
## Testing
- [X] Added/updated automated tests
- [X] QA'd all new/changed functionality manually
- [X] ran `fleetctl gitops` on main and saw a bunch of deprecation
warnings, ran it on this branch and the warnings were gone 💨
- [X] same with `fleetctl generate-gitops`
- [X] ran `fleetctl get` commands and verified that the new URLs and
params were used
- [X] ran `fleetctl apply` commands and verified that the new URLs and
params were used
Fixes#41009
## Summary
- Scope `ValidatePackageIdentifiers` to only run when `$PACKAGE_ID` or
`$UPGRADE_CODE` template variables are present in the uninstall script
- Move `dmg`/`zip` early return before validation
- Switch from ASCII allowlist to shell metacharacter denylist, allowing
legitimate non-ASCII product names (e.g., `®`, parens) while still
blocking injection characters
## Test plan
- [x] Added unit tests for conditional validation (non-ASCII IDs
with/without template vars, dmg/zip bypass, upgrade code scoping)
- [x] Existing input tests still pass
- [x] Winget ingester tests unaffected
---------
Co-authored-by: Ian Littman <iansltx@gmail.com>
Resolves#40857.
The scheduled CI runs (with -race enabled) were failing due to a data
race in ErrorWithUUID.UUID(). The race occurred between:
- HTTP response encoding calling UUID() to lazily initialize the uuid
field
- Error store background goroutine calling Error() via value-receiver
methods, which copies the struct (including the uuid field) concurrently
- Logging calls
Fix:
1. Use sync.Once for thread-safe lazy UUID initialization
2. Change all value-receiver methods on types embedding ErrorWithUUID to
pointer receivers to prevent struct copying that triggers the race
3. Add isNotFoundErr() helper to replace broken errors.Is/errors.As
patterns that relied on value-type error comparisons
From Claude Code Web (ported from my personal fork due to repo access
level required). I've read through the code prior to submitting this PR.
Prompt:
> The scheduled run of .github/workflows/test-go.yaml has had a bunch of
errors in integration tests, starting recently. set up and run the tests
(including race detection) as if you were running in GotHub Actions,
then figure out when the issue was introduced, and what needs to happen
to fix the test errors.
I expect that smoketests and continued during-dev validation of `main`
leading up to 4.83.0 will be sufficient manual testing here.
## Testing
- [x] Added/updated automated tests
- [ ] QA'd all new/changed functionality manually
---------
Co-authored-by: Claude <noreply@anthropic.com>
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
Fixes#34068 Adds automatic retries (up to 3 attempts) for failed software installs from host details, self-service, and setup experience across all installer types.
Resolves#38484. This includes a CI job change to make sure we don't
introduce any more env vars that don't get proxied (and thus turned off
outside `--dev`).
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements)
## Testing
- [x] Added/updated automated tests
Manual QA touched hot paths, but did _not_ manually test every
FLEET_DEV_* environment variable change.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Centralized dev-mode environment management for consistent FLEET_DEV_*
handling and test-friendly overrides.
* Dev-mode allows targeted overrides for certain dev-only configuration
when running with --dev.
* **Chores**
* Migrated environment access to the centralized dev-mode helper across
the codebase.
* Added CI checks to enforce proper usage of FLEET_DEV_* variables.
* **Documentation**
* Added guidance on dev-mode environment variable rules and overrides.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Victor Lyuboslavsky <2685025+getvictor@users.noreply.github.com>
fixes: #32164
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
George Karr
