chore: cleanup code comments

.github/actions/e2e-analyze/README.md

@@ -70,7 +70,7 @@ When the report format changes intentionally, regenerate the snapshots:
 go test ./.github/actions/e2e-analyze/ -update
 ```
 
-## Per-failure bundles (Stage 2)
+## Per-failure bundles
 
 After parsing, the action materialises a bundle directory per failure
 under `<log-dir>/analysis/NN-<TestName>/`. The report links to it from

.github/actions/e2e-analyze/analyze.go

@@ -1,6 +1,6 @@
 // Package main implements the e2e-analyze GitHub Action.
 //
-// Stage 1: cluster failures from go-junit-report's junit-*.xml by error signature,
+// Cluster failures from go-junit-report's junit-*.xml by error signature,
 // pick the earliest non-cascade failure per suite as the likely root cause,
 // and render a human-readable report.
 //
@@ -80,7 +80,7 @@ type Failure struct {
 	// slicing a plausible-looking window around the wrong minute.
 	SuiteSpan time.Duration
 	// BundlePath is the relative path (from log-dir) to the per-failure
-	// bundle written by Stage 2's WriteBundles. Empty when bundles were
+	// bundle written by WriteBundles. Empty when bundles were
 	// not generated (e.g. unit tests that call Render directly).
 	BundlePath string
 }
@@ -552,7 +552,7 @@ func ParseFile(path string, data []byte) (SuiteReport, error) {
 	return rep, nil
 }
 
-// Render writes the markdown report (Stage 1).
+// Render writes the markdown report.
 func Render(r Report) string {
 	var b strings.Builder
 	fmt.Fprintf(&b, "## %s\n\n", r.Title)

.github/actions/e2e-analyze/bundle.go

@@ -17,9 +17,9 @@ import (
 )
 
 // bundlePad is the per-failure window padding either side of the
-// [StartAt, StartAt+Duration] testcase span. Five seconds matches the
-// plan doc: long enough to catch the daemon's pre-test setup and
-// post-test cleanup chatter, short enough to keep the slice focused.
+// [StartAt, StartAt+Duration] testcase span. Five seconds catches the
+// daemon's pre-test setup and post-test cleanup chatter while keeping
+// the slice focused.
 const bundlePad = 5 * time.Second
 
 // nameSanitiseRe replaces characters unsafe in directory names with `_`.

.github/actions/e2e-analyze/main.go

@@ -45,7 +45,7 @@ func main() {
 		report.Suites = append(report.Suites, sr)
 	}
 
-	// Stage 2: prefer authoritative per-suite start times written by the
+	// Prefer authoritative per-suite start times written by the
 	// workflow over junit-report's conversion-time `timestamp` attribute,
 	// then materialise per-failure bundles so Render can link them.
 	if logDir != "" {

.github/actions/e2e-analyze/slice.go

@@ -1,6 +1,6 @@
 package main
 
-// Stage 2: per-failure log slicers used by bundle.go to materialise time-window /
+// Per-failure log slicers used by bundle.go to materialise time-window /
 // testcase-window log excerpts next to each named failure.
 
 import (

.github/workflows/e2e-ddil.yml

@@ -1,8 +1,7 @@
 name: E2E DDIL
 
-# Phase 1 trigger: workflow_dispatch only. Nightly cron + on-main-merge
-# land in mulga-siv-5.8 once ≥4 scenarios are asserting. See
-# docs/development/improvements/ddil-e2e-test-harness.md §4 / §Dependencies.
+# This workflow is manually triggered while the DDIL scenarios are being
+# validated.
 on:
   workflow_dispatch:
     inputs:

.github/workflows/e2e-suites.yml

@@ -1,7 +1,5 @@
-# E2E (per-suite blocks) — EXPERIMENTAL, workflow_dispatch only.
-#
-# Restructures the looped e2e.yml into env-grouped job blocks so each suite
-# surfaces as its own node in the Actions graph (mulga-siv-283):
+# E2E per-suite blocks. Each suite appears as its own node in the Actions
+# graph:
 #
 #   single_provision ─┬─ single_suite (matrix: the single-node-eligible suites)
 #                     └─ single_teardown (if: always)

.github/workflows/gpu-e2e.yml

@@ -1,6 +1,6 @@
 # GPU E2E — drives wattle's persistent, shared-host Spinifex stack over SSH to
 # exercise GPU passthrough (raw EC2, ECS task GPU exposure, EKS pod GPU
-# exposure). See docs/development/feature/wattle-deploy-and-ci.md (Part B).
+# exposure).
 #
 # ---------------------------------------------------------------------------
 # ONE-TIME HOST-SIDE SETUP — NOT part of this repo, a human must do this:
@@ -32,7 +32,6 @@
 #      to skip cleanly (job succeeds, all work steps skipped) instead of
 #      queuing behind manual work indefinitely.
 #
-#   Runbook: docs/development/feature/wattle-deploy-and-ci.md (Part B).
 # ---------------------------------------------------------------------------
 
 name: GPU E2E
@@ -236,9 +235,8 @@ jobs:
       # (SPINIFEX_E2E=1 SPINIFEX_MODE=single, ./<suite>.test -test.v
       # -test.timeout=NNm), plus SPINIFEX_WAN_IP: wattle's ~24 bridges make
       # net.InterfaceAddrs() enumeration pick a co-tenant bridge instead of
-      # the advertised gateway (mulga-7suds), so the override is required
-      # here (proper fix tracked separately: tests/e2e/harness/env.go should
-      # prefer spinifex.toml's advertised bridge).
+      # the advertised gateway, so the override is required here. The harness
+      # should prefer spinifex.toml's advertised bridge.
       - name: Run GPU e2e suite on wattle
         id: e2e
         if: steps.gate.outputs.paused != 'true'

.github/workflows/segscan-oracle.yml

@@ -8,8 +8,8 @@
 # every PR; adding a second-repo checkout there would make every spinifex PR
 # depend on mulga main being healthy and pay that checkout cost on every run,
 # for one storage-accounting oracle test. Mirrors e2e.yml's push-only trigger
-# (mulga-siv-283 precedent): this never gates a PR before merge, matching how
-# e2e.yml's own "front gate" job already works.
+# This never gates a PR before merge, matching how e2e.yml's own
+# "front gate" job works.
 name: Segscan Storage Oracle
 
 on: