Qualify the shared cadence and prepare schedule cutover

.github/workflows/deno-deno-cargo-product.yml

@@ -1,7 +1,6 @@
 ---
 name: 'deno: Deno Cargo product run'
-run-name: "${{ format('Deno Cargo fresh | {0} | CLI {1}', inputs.cache_profile ||
-  'cargo-product', inputs.cli_version || 'Action default') }}"
+run-name: "Deno Cargo / ${{ inputs.cache_profile || 'cargo-product' }} | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} fresh | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   actions: read
   contents: read

.github/workflows/deno-deno-cargo-rolling-chain.yml

@@ -1,8 +1,6 @@
 ---
 name: 'deno: Deno Cargo rolling run'
-run-name: "${{ github.event_name == 'push' && 'Deno Cargo rolling pinned source update'
-  || format('Deno Cargo rolling {0} | {1} to {2}', inputs.cache_profile || 'cargo-product',
-  inputs.base_sha, inputs.head_sha) }}"
+run-name: "Deno Cargo / ${{ inputs.cache_profile || 'cargo-product' }} | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} rolling | ${{ inputs.base_sha }} to ${{ inputs.head_sha }}"
 permissions:
   actions: read
   contents: read

.github/workflows/grpc-grpc-bazel-benchmark.yml

@@ -1,8 +1,6 @@
 ---
 name: 'grpc: gRPC Bazel Benchmark'
-run-name: "${{ format('Benchmark {0} | {1}', inputs.cli_version != '' && 'Canary'
-  || 'Rolling', github.event_name == 'push' && 'Upstream update' || format('Manual
-  dispatch ({0})', github.ref_name)) }}"
+run-name: "gRPC Bazel | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} rolling | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   contents: read
   actions: read

.github/workflows/grpc-grpc-bazel-fresh-benchmark.yml

@@ -1,10 +1,6 @@
 ---
 name: 'grpc: gRPC Bazel Fresh'
-run-name: "${{ format('Benchmark {0} | {1}', github.event_name == 'pull_request' &&
-  'Validation' || inputs.cli_version != '' && 'Canary' || 'Fresh', github.event_name
-  == 'schedule' && 'Weekly' || github.event_name == 'pull_request' && format('PR #{0}',
-  github.event.pull_request.number) || format('Manual dispatch ({0})', github.ref_name))
-  }}"
+run-name: "gRPC Bazel | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} fresh | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   contents: read
   actions: read

.github/workflows/native-fresh-benchmark.yml

@@ -1,7 +1,6 @@
 ---
 name: Native comparison fresh
-run-name: "${{ inputs.case_id }} ${{ inputs.variant }} fresh ${{ inputs.series_id }} sample ${{ inputs.sample
-  }}"
+run-name: "${{ inputs.case_id }}${{ inputs.variant != '' && format(' / {0}', inputs.variant) || '' }} | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} fresh | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   contents: read
   actions: read

.github/workflows/native-rolling-benchmark.yml

@@ -1,7 +1,6 @@
 ---
 name: Native comparison rolling
-run-name: "${{ inputs.case_id }} ${{ inputs.variant }} rolling ${{ inputs.series_id
-  }} sample ${{ inputs.sample }}"
+run-name: "${{ inputs.case_id }}${{ inputs.variant != '' && format(' / {0}', inputs.variant) || '' }} | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} rolling | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   contents: read
   actions: read

.github/workflows/nix-fresh-benchmark.yml

@@ -1,7 +1,6 @@
 ---
 name: Nix comparison fresh
-run-name: "${{ inputs.case_id }} fresh ${{ inputs.series_id }} sample ${{ inputs.sample
-  }}"
+run-name: "${{ inputs.case_id }} | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} fresh | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   contents: read
   actions: read

.github/workflows/nix-rolling-benchmark.yml

@@ -1,7 +1,6 @@
 ---
 name: Nix comparison rolling
-run-name: "${{ inputs.case_id }} rolling ${{ inputs.series_id }} sample ${{ inputs.sample
-  }}"
+run-name: "${{ inputs.case_id }} | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} rolling | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   contents: read
   actions: read

.github/workflows/obs-rolling-benchmark.yml

@@ -1,7 +1,6 @@
 ---
 name: OBS Studio comparison rolling
-run-name: obs-studio ${{ inputs.cache_tool }} rolling ${{ inputs.series_id }} sample
-  ${{ inputs.sample }}
+run-name: "obs-studio / ${{ inputs.cache_tool || 'all' }} | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} rolling | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 'on':
   workflow_dispatch:
     inputs:

.github/workflows/obs-studio-obs-actions-cache.yml

@@ -1,9 +1,6 @@
 ---
 name: 'obs-studio: OBS Studio Actions Cache benchmark'
-run-name: "${{ github.event_name == 'pull_request' && format('OBS validation | PR
-  #{0}', github.event.pull_request.number) || github.event_name == 'push' && 'OBS
-  rolling | Upstream update' || format('OBS Actions Cache | Manual dispatch ({0})',
-  github.ref_name) }}"
+run-name: "obs-studio / ${{ inputs.cache_tool || 'all' }} / Actions Cache | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} fresh | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   actions: read
   contents: read

.github/workflows/obs-studio-obs-boringcache.yml

@@ -1,10 +1,6 @@
 ---
 name: 'obs-studio: OBS Studio BoringCache benchmark'
-run-name: "${{ github.event_name == 'pull_request' && format('OBS validation | PR
-  #{0}', github.event.pull_request.number) || inputs.cli_version != '' && format('OBS
-  canary | CLI {0}', inputs.cli_version) || github.event_name == 'push' && 'OBS rolling
-  | Upstream update' || format('OBS BoringCache | Manual dispatch ({0})', github.ref_name)
-  }}"
+run-name: "obs-studio / ${{ inputs.cache_tool || 'all' }} / BoringCache | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} fresh | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   actions: read
   contents: read

.github/workflows/reapi-fresh-benchmark.yml

@@ -1,7 +1,6 @@
 ---
 name: REAPI comparison fresh
-run-name: "${{ inputs.case_id }} fresh ${{ inputs.series_id }} sample ${{ inputs.sample
-  }}"
+run-name: "${{ inputs.case_id }} | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} fresh | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   contents: read
   actions: read

.github/workflows/reapi-rolling-benchmark.yml

@@ -1,7 +1,6 @@
 ---
 name: REAPI comparison rolling
-run-name: "${{ inputs.case_id }} rolling ${{ inputs.series_id }} sample ${{ inputs.sample
-  }}"
+run-name: "${{ inputs.case_id }} | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} rolling | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   contents: read
   actions: read

.github/workflows/source-case.yml

@@ -10,7 +10,7 @@ permissions:
   actions: read
 jobs:
   previous:
-    name: '${{ inputs.case_id }} previous rolling run'
+    name: Reconcile previous runs
     runs-on: ubuntu-24.04
     timeout-minutes: 10
     permissions:
@@ -48,7 +48,7 @@ jobs:
           if-no-files-found: ignore
           retention-days: 90
   inspect:
-    name: '${{ inputs.case_id }} source check'
+    name: Check source
     needs: previous
     if: needs.previous.outputs.proceed == 'true'
     runs-on: ubuntu-24.04
@@ -87,7 +87,7 @@ jobs:
           if-no-files-found: error
           retention-days: 90
   publish:
-    name: '${{ inputs.case_id }} queue rolling runs'
+    name: Prepare rolling runs
     needs: [previous, inspect]
     if: needs.inspect.outputs.changed == 'true'
     runs-on: ubuntu-24.04

.github/workflows/zed-zed-cargo-product.yml

@@ -1,6 +1,6 @@
 ---
 name: 'zed: Zed Cargo layer matrix'
-run-name: "${{ format('Zed Cargo layers | CLI {0}', inputs.cli_version || 'Action default') }}"
+run-name: "Zed Cargo layers | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} fresh | ${{ inputs.series_id || github.ref_name }} / sample ${{ inputs.sample || '1' }}"
 permissions:
   actions: read
   contents: read

.github/workflows/zed-zed-cargo-rolling-auto.yml

@@ -1,7 +1,6 @@
 ---
 name: 'zed: Zed Cargo rolling run'
-run-name: "${{ format('Zed Cargo rolling {0} to {1}', inputs.base_sha, inputs.head_sha)
-  }}"
+run-name: "Zed Cargo | ${{ startsWith(inputs.cli_version, 'vcli-canary-') && 'Canary' || 'Stable' }} rolling | ${{ inputs.base_sha }} to ${{ inputs.head_sha }}"
 permissions:
   actions: read
   checks: read

cases/obs-studio/payload/scripts/install-ccache.sh

@@ -1,9 +1,9 @@
 #!/usr/bin/env bash
 set -euo pipefail
 
-ccache_version=4.14
+ccache_version=4.14.1
 ccache_archive="ccache-${ccache_version}-linux-x86_64-glibc.tar.gz"
-ccache_sha256=c64760b0b85ba86068f4cd162dc42e2dc39c6f46b0cb8c1990dfccbec7a1fed0
+ccache_sha256=67f8b4cd2620b0d7bc9e834b3baea76bcdca4f2f3f47fe658c343fee58f816fa
 ccache_url="https://github.com/ccache/ccache/releases/download/v${ccache_version}/${ccache_archive}"
 
 storage_version=0.10

cases/obs-studio/payload/scripts/write_phase_result.rb

@@ -15,6 +15,7 @@
 raise "Phase must be base or rolling" unless %w[base rolling].include?(phase)
 raise "Continuous observations require the rolling phase" if args["continuous"] && phase != "rolling"
 cache_hit = args["continuous"] ? args["cache_hit"] : (phase == "base" ? "false" : "true")
+cache_hit = nil if cache_hit == ""
 raise "Cache hit must be true or false" if cache_hit && !%w[true false].include?(cache_hit)
 build_seconds = Float(args.fetch("build_seconds"))
 raise "Build time must be finite and nonnegative" unless build_seconds.finite? && build_seconds >= 0

docs/cadence-cutover.md

@@ -0,0 +1,45 @@
+# Cadence cutover
+
+The new workflows are merged, but central activation is off. The user deferred
+activation until a compatible stable CLI is released. Historical cron triggers
+remain active. Do not apply the prepared patches or set the activation variable
+before the release and qualification gates in [cadence.md](cadence.md) pass.
+
+The [October 5 inventory](../migration/cadence-schedule-inventory-2026-10-05.json)
+records 48 scheduled workflows in the 17 historical repositories represented by
+the maintained suite. It records the default branch, observed head, workflow
+blob SHA, cron expressions and remaining entrypoints. The seven snapshot cases
+have no replaced BoringCache repository schedules.
+
+[Prepared patches](../migration/cadence-cutover-patches) remove only each
+inventoried workflow's `schedule` event. Each edited document was parsed and
+compared with the original after removing that event; manual and PR events, jobs,
+permissions and other workflow behavior remain equivalent. These patches have
+not been applied. Unselected repositories, including Docker corpus, are outside
+this cutover.
+
+When release and qualification are complete:
+
+1. Resolve the exact stable CLI and run the full 31-target dry run on `main`.
+   Require the REAPI capability check to pass without omissions or substitution
+   of a canary. Confirm hosted checks and the retained qualification records.
+2. Refresh the historical inventory. Review any changed workflow blobs before
+   applying their patches. Keep central activation off while removing all 48
+   historical cron triggers. Preserve their manual and PR entrypoints.
+3. Query every historical repository for queued, pending and running scheduled
+   or source-dispatched work. Allow those runs and downstream requests to finish;
+   retain their run IDs. Recheck that no replaced cron triggers remain. Do not
+   cancel unrelated runs or treat a stopped scheduler as a drained queue.
+4. Set `BENCHMARK_CADENCE_ACTIVE=true` in `boringcache/benchmarks`. This transfers
+   source checks, weekly/nightly dispatch and canary monitoring to the central
+   suite. The first hourly source check should produce per-case observations;
+   changed cases should retain intent and requested run IDs independently.
+5. Verify the first complete source cycle and the first weekly and nightly
+   dispatches. Compare each receipt with the 24-case / 31-fresh-target inventory,
+   then inspect leaf outcomes and comparison reports. Request acceptance alone
+   is not build completion.
+
+If central ownership must be stopped, unset the activation variable first and
+inspect already requested central runs. Restore historical cron triggers only
+after central requests have drained, so both owners do not request the same
+work. Retain source receipts and observations during either transition.

docs/cadence-qualification-2026-10-05.md

@@ -0,0 +1,66 @@
+# Cadence qualification — October 5, 2026
+
+All seventeen rolling targets have passed a seed run and an actual source
+advancement with both providers. Zed Nix also passed both fresh samples. This is
+an operational correctness review; these single-sample pairs do not establish
+comparative performance. All runs use `vcli-canary-7a5b27146ebe`.
+
+The [machine-readable inventory](../migration/cadence-qualification-2026-10-05.json)
+records frozen plans, exact sources, cache scopes and selected run IDs. Earlier
+failed and cancelled cohorts remain in `results/`; corrected cohorts have new
+plans. The [evidence inventory](../migration/evidence.json) links archives that
+were downloaded again and verified after publication.
+
+Large archives use a JSON index as `bundle_url`. Its `sha256` and `bundle_bytes`
+describe that index; `archive_sha256` and `archive_bytes` describe the reconstructed
+tar.gz file. Download and verify each listed part, concatenate the parts in order,
+verify the resulting archive, then extract it. The hosted preservation workflow
+performs those checks and verifies every exported file before recording success.
+
+| Case / variant | Seed | Changed source | Result |
+| --- | --- | --- | --- |
+| immich | [37321739971](https://github.com/boringcache/benchmarks/actions/runs/37321739971) | [37323121773](https://github.com/boringcache/benchmarks/actions/runs/37323121773) | Passed |
+| mastodon / server | [37321744434](https://github.com/boringcache/benchmarks/actions/runs/37321744434) | [37323126507](https://github.com/boringcache/benchmarks/actions/runs/37323126507) | Passed |
+| mastodon / streaming | [37321748690](https://github.com/boringcache/benchmarks/actions/runs/37321748690) | [37323131290](https://github.com/boringcache/benchmarks/actions/runs/37323131290) | Passed |
+| n8n / distroless | [37321753318](https://github.com/boringcache/benchmarks/actions/runs/37321753318) | [37323140444](https://github.com/boringcache/benchmarks/actions/runs/37323140444) | Passed |
+| n8n / docker | [37321758035](https://github.com/boringcache/benchmarks/actions/runs/37321758035) | [37323145174](https://github.com/boringcache/benchmarks/actions/runs/37323145174) | Passed |
+| n8n / runners | [37321762921](https://github.com/boringcache/benchmarks/actions/runs/37321762921) | [37323150431](https://github.com/boringcache/benchmarks/actions/runs/37323150431) | Passed |
+| n8n / turbo | [37321767409](https://github.com/boringcache/benchmarks/actions/runs/37321767409) | [37323155106](https://github.com/boringcache/benchmarks/actions/runs/37323155106) | Passed |
+| posthog / layers | [37321772493](https://github.com/boringcache/benchmarks/actions/runs/37321772493) | [37323164071](https://github.com/boringcache/benchmarks/actions/runs/37323164071) | Passed |
+| executorch-buck2 | [37324032100](https://github.com/boringcache/benchmarks/actions/runs/37324032100) | [37324204556](https://github.com/boringcache/benchmarks/actions/runs/37324204556) | Passed |
+| gogs-moon | [37325135541](https://github.com/boringcache/benchmarks/actions/runs/37325135541) | [37325282767](https://github.com/boringcache/benchmarks/actions/runs/37325282767) | Passed |
+| opencut-moon | [37325140807](https://github.com/boringcache/benchmarks/actions/runs/37325140807) | [37325287254](https://github.com/boringcache/benchmarks/actions/runs/37325287254) | Passed |
+| stackstorm-pants | [37324025714](https://github.com/boringcache/benchmarks/actions/runs/37324025714) | [37324228460](https://github.com/boringcache/benchmarks/actions/runs/37324228460) | Passed |
+| msgpack-sbt | [37324036688](https://github.com/boringcache/benchmarks/actions/runs/37324036688) | [37324214890](https://github.com/boringcache/benchmarks/actions/runs/37324214890) | Passed |
+| obs-studio / ccache | [37323465858](https://github.com/boringcache/benchmarks/actions/runs/37323465858) | [37324219008](https://github.com/boringcache/benchmarks/actions/runs/37324219008) | Passed |
+| obs-studio / xcode | [37326133177](https://github.com/boringcache/benchmarks/actions/runs/37326133177) | [37326340221](https://github.com/boringcache/benchmarks/actions/runs/37326340221) | Passed |
+| helix-nix | [37322604229](https://github.com/boringcache/benchmarks/actions/runs/37322604229) | [37323117664](https://github.com/boringcache/benchmarks/actions/runs/37323117664) | Passed |
+| zed-nix | [37322636399](https://github.com/boringcache/benchmarks/actions/runs/37322636399) | [37323174641](https://github.com/boringcache/benchmarks/actions/runs/37323174641) | Passed |
+
+The five REAPI comparator archives identify the intended previous run and source
+in `rolling-seed.json`. Both OBS comparator archives likewise identify their
+intended seed in `previous-cache-source.json`. The review checks those identities
+against the frozen plans; a cache flag alone is insufficient.
+
+The first n8n Turbo advancement retained an available BoringCache entry but
+reported zero native hits. A further actual upstream revision in
+[run 37325183094](https://github.com/boringcache/benchmarks/actions/runs/37325183094)
+reported 70 native hits and 2 misses. Both observations remain retained. Changed
+source does not guarantee a hit; absence of a hit must not be reported as reuse.
+
+Helix Nix passed two fresh cold/warm samples and its rolling pair. Zed Nix passed both fresh cold/warm samples and its rolling pair. Each Nix run imports one checksummed
+dependency store into its provider workers and checks the dependency baseline.
+The measured package is excluded from that seed. A changed package derivation
+may require a build even when the rolling provider cache is available.
+
+The shared controller passed a 24-case hosted planning rehearsal, and fresh
+dispatch passed the complete 31-target dry run. The
+[publication rehearsal](../migration/rehearsals/mastodon-publication/review.md)
+passed isolated conditional publication, dispatch, reconciliation and a forced
+concurrent-write conflict. It used the operator token; the first activated cycle
+must verify workflow-token writes on `main`.
+
+Central activation remains off by request until a compatible stable CLI release.
+The [cutover procedure](cadence-cutover.md) preserves manual entrypoints and
+drains historical runs before central ownership begins. The prepared historical
+schedule patches have not been applied.