Since 4edf5f8265, ChatRequest in internal/api/ai_handler.go has no
autonomous_mode field: Go's decoder drops it and HandleChat passes
approval-required mode into the chat service for every /api/ai/chat
request. The browser still serialized autonomous_mode from its send
options, and the eval runner kept a per-step override that sent it too,
so both read as if the client chose the execution mode while the server
ignored it.
Drop the field from AIChatAPI.chat, SendMessageOptions,
ChatMessageRequestContext and the clone, retry and restore helpers, and
Step.AutonomousMode from internal/ai/eval. aiChatStore.context's
autonomousMode stays as a drawer-only flag, so scoped handoffs still show
"Ask first" and "Approval required before any action." Tests now assert
the request carries no autonomous_mode; the backend test proving a stale
client's autonomous_mode:true cannot elevate a request stays.
Contract prose that described autonomous_mode:false as a per-request
override set by the browser now describes the server-set mode and the
drawer disclosure.
Contract prose still described surfaces that no longer exist:
- `frontend-primitives.md` told new route-level copy to extend
`problemResourcePresentation` (deleted in ae65c5e9a2), tied that helper
to `ProblemResourcesTable.tsx` (deleted with the dashboard overview in
c743a231b8), and kept a readability rule for that table and the
dashboard Pulse Brief. No live surface implements the rule; Patrol's
grouping and affected-resource counts are governed in
`patrol-intelligence.md`. The ownership paragraph and the rule are
dropped, and problem-resource joins the list of deleted dashboard
helpers that must not return.
- `unified-resources.md` said the change and correlation summary cards
default Infrastructure resource-filter links for the Patrol page and the
problem-resource dashboard panels. Since the 2026-05-16 drilldown
retirement the cards have no link default and render only in the
resource drawer, which passes no link builder.
- `api-contracts.md` named the dashboard Pulse Brief as a sender of
`autonomous_mode:false` and described the field as a client override.
Since 4edf5f8265, `ChatRequest` has no such field: the server runs every
`/api/ai/chat` exchange in approval-required mode (pinned by
`TestHandleChat_CannotElevateAutonomousModePerRequest`), and the field
is dropped at decode.
The new `subsystem_lookup_test.py` case pins the deleted helper as
unowned, next to the other retired dashboard paths.
Verification: `contract_audit.py`, `registry_audit.py` and
`status_audit.py` pass with `--check`; the retired-path lookup tests
pass (9/9); the full pre-commit battery passed. No frontend or backend
runtime files change.
frontend-primitives.md still told new route-level copy to extend
problemResourcePresentation, deleted in ae65c5e9a2, and kept a paragraph
tying that helper to ProblemResourcesTable.tsx, deleted with the dashboard
overview in c743a231b8. It also kept a readability rule for that table and
the dashboard Pulse Brief. Neither surface exists and no live surface
implements the rule; Patrol's grouping and affected-resource counts are
governed in patrol-intelligence.md. Drop the ownership paragraph and the
rule, and list problem-resource among the deleted dashboard helpers that
must not return.
unified-resources.md still said the change and correlation summary cards
default Infrastructure resource-filter links for the Patrol page and the
problem-resource dashboard panels. Since the 2026-05-16 drilldown
retirement the cards have no link default and render only in the resource
drawer, which passes no link builder, so labels show as plain text.
api-contracts.md named the dashboard Pulse Brief as a sender of
autonomous_mode:false and described that field as a client override. Since
4edf5f8265 ChatRequest has no autonomous_mode field: the server runs every
/api/ai/chat exchange in approval-required mode, as
TestHandleChat_CannotElevateAutonomousModePerRequest pins, and the field
that scoped handoffs still send is dropped at decode. Say so, and stop
calling Patrol finding handoffs stricter than ordinary chat.
The new lookup test pins the deleted helper as unowned, next to the other
retired dashboard paths, so the dropped ownership claim cannot come back
through the registry unnoticed.
buildPatrolAssessmentAssistantHandoff lost its last caller when the Patrol
summary card's Discuss handler was removed in eac7dfe9ef (2026-05-15), and
buildPatrolConfigurationFailureHandoff lost its last caller in 27ba85ddc6
(2026-06-23). Since then only their own tests called them. A TypeScript
compiler import graph rooted at src/index.tsx confirms no reachable file
imports either one. Removing them also drops PatrolInvestigationContextSummary,
their input and output types, and the private assessment, recent-change,
correlation and save-failure helpers that only they used (found with
noUnusedLocals until it reported nothing).
Rollup had already tree-shaken all of it: production builds of the parent
tree and of this tree are byte-identical across all 187 dist files, and the
bundle-size check passes. The presentation helpers the dead code imported
keep their live callers.
Their test cases go with them. Two cases covered branches the live finding
handoff still uses, so they now run through
buildPatrolAssistantFindingHandoffActions and buildPatrolAssistantFindingHandoff:
a typed pending action without a legacy approval id stays approval-bound,
and the finding context qualifies its resource with the record's node and
truncates long lines. v8 coverage of the surviving model code matches the
pre-removal baseline except one defensive guard the live caller cannot
reach. The raw-source guards pinned none of the removed code.
The Assistant drawer still restores saved patrol_assessment and
patrol_configuration_failure sessions, so that compatibility path stays.
The contracts described both handoffs as live producers, and described a
Patrol summary card, assessment strip, assessment readout, score chip and
header trust line that the page no longer renders. patrol-intelligence,
ai-runtime, api-contracts, frontend-primitives, security-privacy,
storage-recovery, agent-lifecycle and unified-resources now say what the
code does: those handoffs survive only as saved sessions,
the drawer, session list and API keep restoring them, rules for a returning
summary card, strip or readout are conditional, and handoff_actions and
adjacent-boundary clauses name the finding and attention handoffs that send
them today. The run-history handoff clauses also claimed the browser sends a
[Patrol Run Context] block and scoped resources; it sends the briefing and
run identity, and internal/ai/patrol_assistant_handoff.go rebuilds the rest
from the run ID.
docs/AI.md (and its public mirror) told users the legacy operational score
sits in a collapsed supporting-context section. That section and the score
are gone; the section now points at Activity > Finding options and history,
where check history lives. ai_runtime_docs_policy_test.py, its registered
proof, now pins that location and rejects the retired claim. The same file
had failed on main since 2789a15fe5 reworded the Patrol mode API sentence to
"The API retains `autonomy_level` for compatibility" without updating its
pin; the pin now matches that wording.
PlatformResourceTable was the only renderer of UnifiedResourceTable, and
nothing imports it. Every platform page (Proxmox, Docker, Kubernetes,
TrueNAS, vSphere, Machines) renders its own native tables, and
/infrastructure only shows the opening-workspace placeholder. A
production build of main already tree-shook the whole cluster (none of
its strings ship), yet it still carried 11 source files, 7 test files,
registry ownership in two subsystems and contract rules for a table
nobody sees.
Removed: PlatformResourceTable and its PLATFORM_STATUS_FILTER_OPTIONS;
UnifiedResourceTable with its host, service, PBS and PMG cards, source
badge cell, state hook, viewport sync and models; and
ClusterDeployBanner, which only the host card rendered (the live cluster
deploy flow is the Settings ClusterAgentDeployDialog). Their tests go
with them.
Coverage for live code those tests also exercised is kept: the detail
mapper no-fetch boundary and the online boolean sensor case move to
resourceDetailMappers.test.ts, the ResourceFacetSummary render cases
move to their own test file, and both files become registered proofs for
the policies that govern their subjects. Guardrail tests, the shared
template registry, registry.json, status.json evidence and the
performance, unified-resources, frontend-primitives and cloud-paid
contracts drop the deleted paths and the rules that only governed them.
Rollup used to name the large shared chunk after sharedPlatformPage; it
now names it MetricBar, so the bundle baseline entry is re-keyed at its
measured size.
Since 4 October 2026 releases are mechanical: the release train cuts the
next minor from main 14 days after the last stable (or earlier on
request), soaks the release candidate on Preview for 24 hours, and
promotes it unchanged unless an open release-blocker issue stops it.
The shipped release docs, the promotion policy, the checklist, the source
of truth, several governance pages and two workflow comments still
described the retired model: release lines, steward judgment, 72-hour
soaks, a nightly rehearsal and patch releases scheduled by severity.
State the train's rules once, in the Release Train section of the
promotion policy, checked against the train itself, and point the rest
at it. Remove the unused internal copy of the promotion resolver, which
still enforced the old 72-hour soak.
An amber vSphere datastore or network gave no reason on its row: the
vCenter alarm behind it sat one click away in the drawer or one tab
away on the Overview. The alarm text itself led with boilerplate
("Network network-302 has VMware alarm ... (yellow)"), so the Overview
truncated the part that mattered.
A vCenter incident summary is now what vCenter flagged: the alarm's
own name, or "vCenter health is <colour>". Every surface that shows
it already names the resource and severity. Alert cards, notification
templates, and Patrol titles carry the resource name. Patrol's seed
context now names the resource on each alert line, and Patrol titles
key off the incident code instead of parsing the wording, while still
reading records persisted in the old form.
The alert card used to find vSphere by looking for "vmware" in the
message. It now routes on the incident provider, which also sends ESXi
host alarms to vSphere instead of to Machines, where hosts are not
listed. The card's own test becomes a registered alerts proof.
The Datastores and Networks tables gain the exception-first Health
column that TrueNAS storage uses. The cell moves into a shared
PlatformIssueReason, so the two platforms cannot drift; TrueNAS
renders the same. A row's dot and its reason now come from one
classification. Hosts shows only from a 72rem container, where it
reads instead of truncating, and the tablet band splits columns by
weight.
The non-required **Benchmarks** check has failed on every non-superseded
PR since about 15:00 UTC on 5 October. Examples are #2521, #2520, #2524,
#2525, #2527, #2530 and #2532, each ending with `Paired base and
candidate benchmarks (pull requests)` timed out after 25 minutes. The
paired step runs 10 base and 10 candidate samples, so each second added
per sample costs 20 seconds of step time.
### What made it slower
Paired step duration across PR runs, with per-package seconds per sample
taken from the `ok` lines in the job logs:
| Period | Paired step | `internal/unifiedresources` |
`internal/monitoring` |
|---|---|---|---|
| 24 Sep – 5 Oct 08:00 | 16–18 min | 7.9 s | 4.5 s |
| after #2486 (08:11) | 20.5–20.9 min | 16.2 s | 5.0 s |
| after #2518 (15:01) | killed at 25 min | 38.2 s | 9.9 s |
- **#2518** added `BenchmarkRegistryProjectionSnapshot`. Its `cold` mode
rebuilt the registry between `b.StopTimer` and `b.StartTimer` on every
iteration. Go picks `b.N` from timed work only, so the empty-registry
cases ran hundreds of thousands of untimed ingests. Per sample,
`/0/cold/point-targets` took 10.3 s and `/0/cold/bulk-targets` 5.9 s at
a 100 ms benchtime. That is the cause of the timeouts.
- **#2486** added 46 owner-index and metadata sub-benchmarks, about +8 s
per sample. That is legitimate coverage. #2518's
`BenchmarkBroadcastMetricsSnapshot` and #2527 added about +6 s more.
- `-count`, `-benchtime` and the package list did not change.
### Fix
1. **Root cause.** `cold` now invalidates the registry's derived state
(canonical metadata, typed views, source-target index) instead of
rebuilding the registry. The refresh never reads its own output, so this
repeats a fresh ingest's first read: at 1000 VMs, 40011 vs 40014
allocs/op and about 5.09 MB/op, the same as before. Locally the 24-case
family now runs in 4.9 s; before, its 6 cold cases alone took 33.8 s.
2. **Scope.** Paired PR runs now sample only packages whose in-module
dependency closure (`go list -deps -test`) holds a changed file. A
`_test.go` file selects only its own package. Changes to
`go.mod`/`go.sum`, dirty or non-Git trees, or an unreadable graph select
every package. Applied to today's failures: #2521, #2524 and #2525
sample nothing; #2530 samples `internal/api`; #2532 samples api and
monitoring; #2520 samples 5 packages; #2527 samples 7. The metadata
records `selection=` and `packages=`.
3. **Budget.** A module or shared-model change still samples nearly
everything, about 24 minutes including the ~2.7 min warm-up. The step
timeout is now a measured 30 minutes (job stays at 35).
`PULSE_BENCH_BUDGET_SECONDS=1680` makes the script project its total
after each round. Once the projection reaches 28 minutes it stops and
prints each package's mean seconds per sample, which replaces an
unexplained timeout. `TestPairedBenchmarksStopInsideTheirBudget` keeps
budget < step timeout < job timeout.
4. **Store-backed chart benchmarks** (second commit). Go calls a
benchmark function again for each `b.N` it tries.
`HandleWorkloadCharts_StoreBacked` and
`HandleWorkloadsSummaryCharts_StoreBacked` each seeded 84,000 points
into a fresh SQLite store (about 0.6 s) to time a request of 0.2 ms or
less, two or three times per sample. In CI that was 3.5 s and 3.6 s of
internal/api's 11.5 s per sample. The timed loop now runs in a
`range=4h` sub-benchmark, so the parent seeds once per run. Data,
request and bytes/allocs per op are unchanged, and locally the pair
falls from about 6 s to 1–1.5 s. The names gain `/range=4h`, so this
PR's comparison reports them unmatched.
Unpaired branch runs still sample every package. With the cold fix,
their 15-minute step (13–14.2 min today) regains about 3 minutes of
headroom.
### Verification
- `scripts/tests/test-ci-benchmarks.sh` covers no-op, own-test-only,
shared-dependency, unreadable-graph, dirty-tree, module-file,
budget-exceeded, in-budget and invalid-budget cases. Five mutations of
the selection and budget logic are each caught.
- On the real package graph, selection maps `internal/models` to 7
packages, `internal/monitoring` to api+monitoring, an `internal/api`
embed to api, and frontend or docs paths to none.
- End to end with real Go on two clean clones (`origin/main` vs this
change), only `./internal/unifiedresources/` was sampled. Its per-sample
time went from 53.9 s to 26.7 s on a loaded Mac.
- `go test ./scripts/installtests -run
'TestPairedBenchmarksStopInsideTheirBudget|TestBenchmarkQualificationRetainsProvenance'`,
plus the script-reference and doc-drift suites.
This PR's own Benchmarks run should sample only
`internal/unifiedresources` and `internal/api`, because both changes are
those packages' own test files.
Since 15:00 UTC on 5 October every paired benchmark run that was not
superseded has hit its 25-minute step timeout. Each step runs ten base
and ten candidate samples, so per-sample growth is multiplied by twenty.
BenchmarkRegistryProjectionSnapshot (#2518) rebuilt a registry behind
StopTimer on every cold iteration. Go sizes b.N from timed work only, so
the empty-registry cases ran hundreds of thousands of untimed ingests
and took 16 s of a 100 ms benchtime. With #2486's new owner-index and
metadata benchmarks, internal/unifiedresources went from 8 s to 38 s per
sample. Cold now invalidates the registry's derived state instead. The
refresh never reads its own output, so this repeats a fresh ingest's
first read with the same allocations, and the family drops from 22 s to
about 5 s per sample.
Paired runs also measured every package even when both trees compiled
identical code, as on frontend-only changes. They now sample only
packages whose in-module dependency closure the change reaches. Module
files, dirty or non-Git trees and an unreadable graph still select
everything.
A module or shared-model change still samples nearly every package,
about 24 minutes with warm-up. The step therefore gets a measured 30
minutes, and the script stops at a 28-minute projected budget with
per-package costs instead of meeting a bare timeout.
Promote v6.5.0-rc.1 to stable after its 24-hour soak. Only release metadata changes.
Contract-Neutral: Release-only v6.5.0 version, release notes and distribution metadata; no public install or runtime contract changes.
Change-source: pulse-maintainer
Preserve the exact documentation and regression-test candidate; native browser follow-through and the missing UI remain owned separately.
Change-source: pulse-maintainer
Explain the missing v6.4.5 undo control and the existing scoped API. Verify exact selection, confirmation, browser session and CSRF, failure handling and readback without copying credentials or deleting unrelated rules.
Contract-Neutral: Documentation and regression-proof coverage of existing Patrol suppression APIs only; no runtime, route, authority or subsystem contract changes.
Change-source: pulse-maintainer
## Problem
An agent installed by the cluster deploy flow starts with `--enroll` and
a one-time bootstrap token. Only the host module exchanged that token
for the runtime token. Everything else `main.go` builds at startup kept
the bootstrap token for the whole first process lifetime:
- the remote config client (initial fetch and the one-minute refresh)
- the updater
- the Docker / Podman and Kubernetes modules and their primary report
targets
The bootstrap token carries only the enrollment scope and the server
removes it on exchange, so each of those failed with 401 until the
service restarted. On a freshly deployed Proxmox node this showed up as
a remote config 401 every minute. Those modules also kept the
pre-enrollment agent ID instead of the server-assigned one.
## Fix
`main.go` now enrolls once, before it builds any authenticated client:
- `hostagent.Enroll` is a standalone entry point over the existing
enrollment loop. Payload, headers, retry and backoff, status handling,
and `runtime.token` / `agent-id` persistence are unchanged.
- It runs after self-test and agent-ID preparation and before the first
remote config fetch. Every module is then built from the runtime token
and the canonical agent ID.
- The ID is persisted to the configured agent-ID file, and the host
module's own enrollment is switched off so there is no second exchange.
- A persisted runtime token is reused without a request or host
discovery. A `403` keeps the configured token as a non-bootstrap
credential. A permanent rejection stops startup before any module is
built.
- Windows is gated off and keeps its previous behaviour. Startup there
runs before the service control dispatcher, and an enrollment retry
could outlast its 30-second deadline. No Windows installer passes
`--enroll`.
- Enrollment and reporting now derive hostname, OS and architecture
through one `resolveHostIdentity`, so the runtime token is bound to the
hostname the agent reports under.
## Tests
- `TestRunEnrollsBeforeBuildingAuthenticatedModules`: enrollment happens
before any request to Pulse. Remote config, the updater, and the host,
Docker and Kubernetes configs and their targets all carry the runtime
token and canonical ID, and the agent-ID file is written. With the new
step disabled, this test fails.
- `TestRunStopsBeforeModulesWhenEnrollmentIsRejected`,
`TestRunWithoutEnrollDoesNotExchangeTokens` and
`TestRunLeavesEnrollmentToTheHostModuleWhenGatedOff`.
- `TestEnroll*` cases for `Enroll`: payload and identity, the
monitoring-only command ceiling, reuse of a persisted token (including
when host discovery fails), the 403 keep, the 409 rejection, and a
missing token.
- `TestEnrollPresentsTheIdentityNewReports` checks that enrollment
presents the hostname, OS and architecture that `New` reports, with
Proxmox detection and the architecture fallback.
- `go test ./cmd/pulse-agent ./internal/hostagent
./internal/remoteconfig` passes. `go vet` is clean for linux, windows
and freebsd.
## Not in this change
- The health listener still starts after enrollment, so `/readyz` is
unreachable while enrollment retries on network errors. The installer
only warns in that case. With the health listener enabled, its
registration check now always sees the runtime token. With it disabled,
the process-only check can still race enrollment.
- Distinguishing a deploy binding mismatch from a non-bootstrap token,
which both return 403.
- Two older durability gaps that this change doesn't make worse. If
writing a custom `--agent-id-file` fails, the next start has no fallback
to the canonical ID saved in the state directory. If persisting
`runtime.token` fails, enrollment still succeeds, and the next restart
is left with the spent bootstrap token.
Every subsystem must declare require_explicit_path_policy_coverage, so
build_verification_requirements never falls back to a rule's default
verification policy; an unmatched runtime file becomes a path-policy gap
instead. The default test_prefixes and exact_files were therefore never
consulted, yet 13 of 16 rules carried them (93 files and one prefix),
and authors kept adding proofs there as if they counted.
92 of the 93 files were already proof on some path policy, and
api-contracts' default prefix is on three of its own policies. The one
orphan, internal/api/router_bench_test.go, benchmarks the api-contracts
router rather than any performance-and-scalability runtime, so no policy
should take it; it stays an owned file only.
The registry drops the lists, registry_audit requires them empty and
points authors at the path policy instead, and the canonical protocol
says so. The protocol also stops promising same-subsystem tests "when
the registry explicitly allows them", which the audit no longer permits.
No verdict changes: build_verification_requirements returns identical
requirements for all 2746 owned runtime paths, and for an uncovered
path, under the old and new registry.
Document host-capacity normalisation, version-qualified rounding and the difference between collected readings and delivered alerts. Keep comparisons on ordinary workload observations rather than forced load, threshold changes or diagnostic reruns.
Contract-Neutral: Existing container CPU semantics and safety guidance only; no runtime, API, threshold, permission or lifecycle contract changes.
Change-source: pulse-maintainer
staged_verification_files_for_requirement() matches
allow_same_subsystem_tests against a subsystem's
owned_prefixes/owned_files, but both commit guards passed it the
impacted entry, which carried neither. Since the v6 import (778a2577b6)
a same-subsystem test never satisfied a requirement, while
subsystem_lookup.py (which passes the registry rule) and the guard's own
BLOCKED message ("same-subsystem test/spec files are also accepted")
said it would.
Honouring the 12 flags would have loosened the guard well past their
apparent intent. Every rule requires explicit path-policy coverage, so
the five rule-level flags are never consulted, and on the seven path
policies the flag reaches every test the subsystem owns: a
mutation-registry change would accept any internal/ai or AI Chat test
(~455) instead of its 4 listed proofs, and a frontend Actions change any
Go test under internal/unifiedresources. Six of the seven already list
their own directory in test_prefixes.
So the strictness stays and the tools now agree on it: the registry sets
the 12 flags to false (what has always been enforced), registry_audit
rejects true and points at test_prefixes/exact_files, and the impacted
entry carries the rule's ownership so the completion and commit-shape
guards read the flag the same way subsystem_lookup does.
No requirement accepts a proof it rejected before. subsystem_lookup
stops reporting matches the guard refused: a new internal/hostagent test
no longer claims three agent-lifecycle policies.
The completion guard's CI mode now compares contract text against the
evaluated commit (#2502), but it still decided which contracts a commit
owed from HEAD's registry and contract index. CI walks a branch with HEAD
at the tip (or at main after a merge), so a later commit could change an
earlier commit's verdict. If a later commit added a Shared Boundaries
reference to a file, an earlier commit that was complete when written got
blocked. If a later commit dropped ownership of a file, an earlier commit
that changed it without its contract passed. The pre-commit hook read
both from the commit's own index, so CI and the hook could still
disagree.
repo_file_io and subsystem_contracts can now read from a revision as well
as the working tree or index. A revision read uses only that commit's
tree and fails closed rather than falling back to the disk. In diff-base
mode the guard reads the registry and contract index from the evaluated
commit, as does the Prettier-only inference that --commit enables.
Replaying CI's per-commit loop over the last 120 commits on main gives
the same verdicts as before.
Expose only the installed GNU gtimeout as timeout through a new runner-owned directory. Keep native BSD utilities on their original PATH so installing the missing dependency cannot silently change the other platform checks. The connected workflow test verifies the symlink identity and that the directory contains no other utilities.
Contract-Neutral: Narrow the native CI dependency PATH and its proof; no shipped installer or runtime contract changes.
Change-source: pulse-maintainer
Both macOS native jobs fail the copied temperature recipes before reading any sensor. Install coreutils into the workflow PATH and verify GNU timeout for every Unix runner. Exercise the actual workflow snippets with real GNU timeout and fail-closed install, missing-tool and unsupported-tool controls; leave the recipes and their TERM/KILL/privacy controls unchanged.
Contract-Neutral: Native CI prerequisite and its fixture coverage only; no installer, runtime, API, permission or diagnostic contract changes.
Change-source: pulse-maintainer
CI and dev-prepush walk a branch's commits with HEAD at the branch tip and
run the completion guard per commit with --diff-base <commit>^. The guard
then compared the parent's contract text against HEAD rather than against
the commit it was evaluating, so every commit was judged by the whole
branch. The pre-commit hook compares HEAD with the staged index, which is
the commit's own parent-to-commit change, so the two disagreed.
PR 2494 hit this: its second commit reverted a Current State paragraph in
performance-and-scalability.md that its first commit wrote. Parent-vs-HEAD
saw no change, and CI blocked a commit the hook had correctly passed. The
same comparison lets a bare runtime commit through when a later commit in
the branch carries the contract update.
When --commit is given, the guard now resolves the diff base and reads the
new contract text from that commit. CI already passes --commit, and
dev-prepush now does too. Without --commit the guard still compares with
HEAD, which is right when the evaluated commit is checked out.
Match QNAP guidance to existing data-volume relocation, create private per-attempt staging and preserve diagnostic logs and identity. Exercise copied commands for preparation and installer failures without an appliance operation.
Contract-Neutral: Clarifies existing installer storage and private staging guidance; no runtime, permission or data-contract change.
Change-source: pulse-maintainer
Preserve the exact reviewed documentation candidate and its retained privacy/query proof; no runtime or storage changes.
Change-source: pulse-maintainer
Reuse the tested API helper, explain point/source and access distinctions, and preserve data before storage tuning. Copied examples now exercise private success/error responses, redirects, incomplete transport and encoded IDs.
Contract-Neutral: Documentation and recipe checks align with existing history API behaviour; no runtime, storage, entitlement or authentication contract changes.
Change-source: pulse-maintainer
Preserve the reviewed candidate commits as the second parent and retain canonical one-shot, redirect, partial-response and disclosure controls in the overlapping API/PBS smoke fixtures. Keep the candidate's canonical PBS guard, scoped private temperature-file assertions and unchanged-verdict audit diagnostics.
Contract-Neutral: Integration of test-only private-help adapters; no runtime, guides, dependencies or frontend source changed.
Change-source: pulse-maintainer
The reviewed phase-aware runner and PBS helper exclude Windows. Apply the same build constraint to their existing Bash temperature consumers so Windows test compilation does not retain unresolved runPBSRecipe references. Linux test content, guides and runtime are unchanged.
Contract-Neutral: Align an existing Bash-only test file with its non-Windows test helper; no runtime, guide, dependency or platform contract change.
Change-source: pulse-maintainer
A bounded synthetic comparison shows the whole-shell watchdog can kill a POST before its own curl deadline or expire before a PBS request starts. Observe each phase privately, preserve curl arguments and exact request limits, cancel owned descendants and retain incomplete pipe results. Exercise delayed preparation, one-shot transport, private responses and every cancellation phase without changing any guide or runtime.
Contract-Neutral: Repair Bash documentation test timing and owned-child observation only; no guide, mirror, runtime, API, authentication, request deadline, dependency or permission contract changes.
Change-source: pulse-maintainer
Bind the new 0600 assertion only to the reviewed temperature cleanup helper rather than impose it on the unchanged public Unified Agent installer recipe. Include the captured canned audit output when its existing one-call assertion fails, preserving verdicts, call counts and timing controls. The complete second proof retains its fail-closed two-call observation without a cause or successful-retry claim.
Contract-Neutral: Test adapter scope and failure diagnostics only; existing guide, audit verdict and timing contracts are unchanged.
Change-source: pulse-maintainer
Keep the shared direct-curl recorder compatible with the other existing guides while binding API requests to its reviewed private helper. Update the temperature selector to the reviewed fail-closed temporary download and retain disable-first TLS and mode checks. The first broad proof caught both adapter faults before publication.
Contract-Neutral: Test-only adapters for existing reviewed guide contracts; no runtime, guide, dependency or frontend change.
Change-source: pulse-maintainer
Repair deterministic governance and script-smoke failures against the reviewed guides without restoring public response output, literal credential commands or clobbering downloads. Exercise the actual shared API helper and independently verify private, unique response files, methods, bodies and hostile curl defaults.
Contract-Neutral: Update test adapters for the existing reviewed credential-private guide contract; no runtime or guide change.
Change-source: pulse-maintainer
Execute the actual API helper and PBS private-file/download handoffs, keeping their response-privacy assertions separate from shared credential-transport fixtures. Verify one-shot requests, exact methods and JSON, retained private responses and failure exits, TLS and preparation. Match the temperature cleanup test to its current non-clobbering download, without changing any guide.
Contract-Neutral: Repair connected documentation fixtures only; no guide, mirror, runtime, API, authentication, dependency or permission contract changes.
Change-source: pulse-maintainer
Replace automatic hardware detection with existing-driver reads, keep SSH forced-key trust intact, and prevent incomplete legacy cleanup downloads from replacing a saved helper. Validate the copied commands against synthetic tools and HTTPS peers.
Contract-Neutral: Existing help and copied-command safety only; sensor collection, authentication, installer interfaces and runtime schemas are unchanged.
Change-source: pulse-maintainer
Bound the existing passive preflight before Bash can discard NULs or accept unlimited configuration output. Preserve producer errors independently, isolate the byte reader from Python startup hooks, and keep the copied helper and mirrored help private and off the guest-agent channel.
Contract-Neutral: Operator diagnostics and existing help only; Proxmox polling, guest-agent coordination, APIs, storage and service behaviour are unchanged.
Change-source: pulse-maintainer
Save every existing reference response privately, including errors and partial POST results. Preserve private header files, reject unsafe preparation and URLs, and stop on redirects or transport errors without retries. Exercise all eight copied GET/POST examples against real synthetic HTTP/TLS peers, including the old response exposure.
Contract-Neutral: Harden existing API documentation recipes and their shipped mirror with connected tests; no runtime, endpoint, authentication, scope, action, dependency or permission contract changes.
Change-source: pulse-maintainer
Retain the guide recipe array before selecting its request slice; the initial exact-source proof rejected a temporary array slice before running any tests.
Contract-Neutral: Correct test compilation only; existing PBS documentation and runtime contracts are unchanged.
Change-source: pulse-maintainer
Fail closed before editing unsafe credential paths, ignore curl configuration and retain diagnostic responses privately. Bound requests, reject redirects and publish the local installer path only after a complete HTTPS 200 download without overwriting existing files. Exercise the copied recipes against synthetic TLS peers and file failures.
Contract-Neutral: Harden existing PBS documentation recipes and their shipped mirror; no runtime, API, permission, installation contract or dependency change.
Change-source: pulse-maintainer
PR2472 build run 37239183545 extracts no wrapped TrueNAS recipes; exercise all three current blocks and retain synthetic HTTP response bodies only in private files, including 201 save and error controls. Native run 37239183481 expects a Linux QGA capability on macOS and Windows; assert non-Linux agents do not advertise it and scope the connected PVE fixture to its actual Linux runtime. Production safety and documentation remain unchanged.
Contract-Neutral: Test-only reconciliation of already-reviewed private-response recipes and existing Linux-only guest-execution capability; no product contract change.
Change-source: pulse-maintainer
Source proofs can supply unrelated input files. Retain recipe mutation controls while reading the exact parent only when its named guide was supplied; unrelated input directories must not become a new full-suite prerequisite.
Contract-Neutral: Test-only optional-input handling; no shipped guidance or product contract change.
Change-source: pulse-maintainer
Make the existing connection guide fail closed before editing credentials, ignore curl configuration, bound each single POST and retain sensitive responses in new private files. Explain actual test/save status and uncertain-save readback without retrying setup as diagnostics.
Contract-Neutral: Documentation-only credential handling and copied-recipe regression tests; no runtime API, TrueNAS polling, authentication or connection-management contract changes.
Change-source: pulse-maintainer
Reviewed exact PI candidate 5e6d5af3b5. Existing guidance and shipped mirror now use private header files, verified HTTPS and bounded read-only checks. Retain source-e1x7r92a installer/vet, source-oypceg0j docs tests and Chromium/WebKit rendering evidence; no native tenant isolation or publication claim.
Change-source: pulse-maintainer
Replace inline token arguments and certificate bypasses with private header files and verified HTTPS. Bound the read-only probes, discard bodies, preserve failed curl exits and distinguish an invalid token from a checked tenant boundary. Keep the shipped guide byte-synchronised. Exercise copied commands with a synthetic TLS peer, private-header canaries, hostile curl configuration and refusal/redirect controls; no customer endpoint is used.
Contract-Neutral: Existing diagnostic guidance and regression coverage only; no runtime, public API, token scope or tenant-isolation contract change.
Change-source: pulse-maintainer
Release candidate cut from main at 2e9404504b. It carries every change merged since stable v6.4.5.
Contract-Neutral: Release-only v6.5.0-rc.1 version, notes and packet metadata generated by pulse-maintainer-release-metadata; no public install or runtime contract changes.
Change-source: pulse-maintainer
## What changes and why
- Keep backup-age checks fast and make failed or partial evaluations
visible through one existing system warning, without clearing retained
age alerts. Indexed latest-observation reads preserve retention, schema
and the read budget. This responds to the recurring missed checks
reported in #2465; synthetic timing is not proof of relief on the
reporter's Pi.
- Defer guest-agent commands after uncertain HTTP results and require
same-target, unambiguous backup-lock evidence before guest reads. Keep
guest coverage truthful when only remembered versions or a parent-node
assignment exist. Preserve the original observation time/source of
retained memory through existing backend projections. These changes
advance #2439 safety, but do not establish native thaw or recovery.
- Attribute Proxmox backups using native connection/node and unique PBS
namespace evidence, retaining ambiguous artifacts rather than assigning
them by substring or display name. Show unreliable PBS capacity as
Unknown/Unavailable, while retaining valid measured zero and readable
phone warnings.
- Restore existing TrueNAS and VMware connection alerts using the
constructed runtime pollers. This does not resolve the distinct
monitoring cause reported in #2382.
- Keep safe report collection and attachment guidance self-contained,
and make manual backup pause/resumption precautions usable. Repair the
Diagnostics integration journey's outdated controls, download assertions
and session capability, without changing production authorization.
## Included source
Published base `61146105c74a5968d877691e02963c41f8e66076` → exact final
head `af73b0e6c618da83c4fc5e702ba464257481851c`.
The previous proposal head `dcde1d0157d5e3c4cd626d6b920a45016a3f0437`
and all reviewed ancestry remain unchanged. This advances the proposal
once after its workflows completed with demonstrated failures; it does
not interrupt a checking head.
Reviewed candidates include:
- Previous batch: guest HTTP `cd1e3ba12f`, runtime alerts `7f5c8a0cf1`,
safe intake `496ba1e73e`, PBS capacity/render sequence `8f09a59996`
through `d2f7e456b3a`.
- Manual pause `3fd44b8325`; native backup ownership `c5e83ceb3a` /
`6198a3ce22`; authoritative locks `b77e6908e6`; truthful guest coverage
`ba3e2d2c1c` / `8f9a9a6a1e` / `f9f53befbb`.
- Backup evaluation/query `5774b6bc4b` / `7a19a06631`; Diagnostics
sequence `ed168c8d7e` through `a3bbcae0c9`; memory provenance
`abee023634` through `80d8ad0f59`; self-contained intake `7c9922f55d`.
- Starting reviewed canonical tip
`8003a4037bf2e6628d47ca6c7da239a11096168e` is retained. Four additive
publication corrections (`745ddc982a` through `79091f7b03`) supply the
report-form contract, connected docsLinks regression and exact
historical completion pair. The first completion's section placement
failed the checker; its substantive-section correction passes. No
accepted commit was amended or recreated.
## Validation and remaining gates
The previous proposal's completed runs remain adverse: Build/Test
`37204956836` failed the documentation drift assertion; E2E
`37204956824` failed Diagnostics shards 3/6/8 (download/obsolete
Run-control timeouts). These warrant the admitted changed-source
repairs, not an unchanged retry.
Retained reviewed evidence includes guest transport/connected/race/build
(`source-5_ta7via`), runtime alert connected proof (`source-s0xmdftf`),
PBS source/render evidence (`source-_9ekxhcy`), manual-pause
source/render (`source-ifgmfzhw`), all-reader authoritative-lock proof
(`source-41amnzmb`), guest coverage's 1,901 connected tests/full
types/lint/build and inspected Chromium/WebKit desktop/phone renders
(`source-3y6kfagt`), backup recovery/alerts/monitoring/races/vet/real
builds (`source-vc1jlsz9`), and memory
models/unified-resource/alerts/monitoring/races/vet/types/real
embed/build (`source-5swnugj9`). Diagnostics retains nine passing
full-app journeys with matching source hashes; catalogue
`source-cyehgbio` discovers 916 test/project identities, including all
nine Diagnostics cases, but is not runtime acceptance. Intake
`source-stznewur` retains two intended parent failures and passing
policy/documentation/mirror checks.
Publication checks: provenance, clean retained ancestry, per-commit
canonical completion/history, registry/contract audits and final
browser-content coverage (nine changed production frontend files) pass.
Earlier proposal-head (`79091f7b`) offline VM proof `source-1fn6chab`
passes 36 docsLinks tests, formatting, 27 intake cases, 211 public
Markdown checks and 62 mirrors, with complete hash-matched output and
guest stop. Its line-index bound is explicit. Earlier `source-yj300mm2`
passed docsLinks but exited 1 when a browser script was mistakenly
invoked in a source VM without Playwright; no browser result is claimed
from that attempt. Earlier local preparation failures and prior
API/embed limitation remain evidence at their own bindings, not green
aggregate results.
Required exact-head CI must run the complete backend/build/frontend,
audit and protected E2E checks, with independent source review, before
landing. Existing sufficient connected/render proof was not repeated
solely for batching. This is source publication, not release
qualification, deployment, native appliance/guest acceptance or
completion of the linked reports. The containing all-main release, Pi
evaluation/failure/recovery, native backup/thaw/covered writes/liveness,
remaining memory drawer consumption and installed whole-panel/alert
results remain separate.
Related to #2465
Related to #2439
Related to #2382
[How Pulse handles
triage](https://github.com/rcourtman/Pulse/blob/main/docs/AI_TRANSPARENCY.md)
## Completed workflow reconciliation and additive smoke repair
The starting proposal head `79091f7b0317807fcc76e36cde52c4ee1cc4c928`
and every reviewed commit remain ancestors of this head. Build/Test
`37219920651` has now completed. All backend partitions, frontend,
governance, security and all eight E2E shards/verdict passed, but script
smoke failed two obsolete whole-guide assertions in
`test_vm_disk_diagnostics.py`; backend build was consequently skipped.
The reviewed manual backup precaution deliberately adds six server/timer
examples. The old passive-helper test counted all seven guide examples
and banned `systemctl` throughout the whole guide. This source mismatch
warrants changed-source repair, not an unchanged retry. Additive
test-only `c4a72995c6bf2465a6540a1b546a9dcd732944af` scopes
passive-helper execution to its named section and admits only the exact
six reviewed manual-pause blocks. Its first local run retained one
failure because a shell comment was mistaken for a heading. Additive
`af73b0e6c618da83c4fc5e702ba464257481851c` parses headings only outside
fenced blocks. All 17 diagnostics/safety controls now pass, including
missing/duplicate headings, shell comments and five prohibited-command
mutations. The passive helper still rejects active guest requests and
service mutations. Production and guidance bytes remain unchanged;
mirrored docs and connected intake checks pass. Both failed
correction/control observations remain recorded, not relabelled.
Final publication provenance, clean retained ancestry, additive
contract/registry audits and the unchanged nine-production-file
browser-content coverage pass. Lightweight confined fixture proof
supplies the full 17-case suite; existing heavy/render proof was not
repeated for this test-only delta. This proposal advances only after the
previous head's workflows completed. Exact new-head CI must rerun its
required checks before landing.
The completed advisory benchmark job is also adverse:
`RefreshPlatformScopes` 294.1 → 325.0 ns (+10.51%) and
`ClassifyResourceSensitivity` 67.68 → 77.20 ns (+14.07%), each p=0.000
with ten samples. Their implementation and benchmark source are
unchanged across the batch; that does not prove noise, intentional cost
or lack of regression. Delivery owns a bounded evidence-based
disposition through the existing lane handoff. No favourable repeat,
threshold change or performance/release approval is claimed.
[How Pulse handles
triage](https://github.com/rcourtman/Pulse/blob/main/docs/AI_TRANSPARENCY.md)
rcourtman