@devrouter/cli 0.0.51 → 0.0.53

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@devrouter/cli",
3
- "version": "0.0.51",
3
+ "version": "0.0.53",
4
4
  "description": "Local dev routing CLI with shared Traefik reverse proxy",
5
5
  "author": "Roland Schlaefli",
6
6
  "homepage": "https://github.com/rschlaefli/devrouter#readme",
@@ -0,0 +1,34 @@
1
+ # Upgrade to devrouter 0.0.52
2
+
3
+ Managed process ownership now handles large environments correctly. Explicit
4
+ `ensure --repair` recovers a retained degraded runtime without deleting its data
5
+ or invoking provider bootstrap. Ordinary `ensure` retains its degraded-state guard.
6
+
7
+ ## Task
8
+
9
+ 1. Install the published CLI with `npm install -g @devrouter/cli@0.0.52`.
10
+ 2. Bump `devrouter.version` in the consumer `.devrouter.yml` to `0.0.52`.
11
+ 3. For an approved exact workspace recovery, inspect status and run
12
+ `devrouter ensure <checkout> --repair --json`. Omit the profile to reuse the
13
+ recorded profile. Preserve required non-secret Compose interpolation settings.
14
+ 4. Stop at ownership or configuration drift. Repair requires retained containers;
15
+ it never creates missing resources or changes the recorded profile.
16
+
17
+ ## Validation
18
+
19
+ Verify installed/configured versions with `devrouter -V --repo <checkout>`.
20
+ Repair accepts a stopped primary only with every project container explicitly
21
+ stopped and no checkout routes. It starts retained IDs, including their existing
22
+ entrypoints, then replays the current repository adapter. Owned processes may
23
+ restart when fingerprints change. Failure can leave owned resources running.
24
+
25
+ Require zero command exit status, ready services/processes, exact active route
26
+ rules and targets, and no managed drift. Check the application separately.
27
+ No live recovery is implied by installing the CLI. No schema migration or manual
28
+ state-file edit is required.
29
+
30
+ ## Report
31
+
32
+ Report installed/configured versions, exact checkout/profile, repair exit status,
33
+ managed state, application check, and remaining blockers. Distinguish source
34
+ checks from live runtime proof. Data deletion remains a separate explicit action.
@@ -0,0 +1,42 @@
1
+ # Upgrade to devrouter 0.0.53
2
+
3
+ Managed process adapters can prepare dependencies synchronously before launching
4
+ the background application. Exact managed Devsy stop can finish retained service
5
+ shutdown after the primary has stopped.
6
+
7
+ ## Task
8
+
9
+ 1. Install the published CLI with `npm install -g @devrouter/cli@0.0.53` and
10
+ verify `devrouter-process --help` advertises `--prepare-command`.
11
+ 2. Bump `devrouter.version` in the consumer `.devrouter.yml` to `0.0.53` when
12
+ updating that repository's release pin. No schema migration is required.
13
+ 3. If dependency preparation currently consumes application readiness time,
14
+ adapt the repository-owned managed adapter to call
15
+ `devrouter-process ensure --prepare-command <command>` before its application
16
+ command. Keep preparation in the foreground; do not daemonize or detach.
17
+ Unchanged owned processes skip preparation. Explicit fingerprints remain
18
+ caller-owned and must encode any required preparation identity.
19
+ 4. For an approved retained-runtime recovery, preserve required Compose
20
+ interpolation settings and use `devrouter stop <checkout>`. A stopped primary
21
+ is not sufficient shutdown evidence: require every project container stopped
22
+ and zero checkout routes before guarded `ensure --repair`.
23
+
24
+ ## Validation
25
+
26
+ Verify installed/configured versions with `devrouter -V --repo <checkout>`.
27
+ Check preparation failure/cancellation, warm reuse, application readiness and
28
+ exact shutdown for the affected consumer. Preparation must finish before its
29
+ readiness deadline begins.
30
+
31
+ Stop validates ownership, configuration and complete container membership on
32
+ every mutation boundary. Changed, missing or unreadable evidence blocks cleanup.
33
+ An original provider failure is still reported after successful residual service
34
+ cleanup; inspect fresh state before choosing another operation. No containers,
35
+ volumes or runtime data are deleted. Repair retains its unchanged-configuration
36
+ guard and never creates missing resources or invokes provider bootstrap.
37
+
38
+ ## Report
39
+
40
+ Report installed/configured versions, exact checkout and retained profile,
41
+ preparation and readiness outcomes, all-container shutdown proof, route count,
42
+ and remaining blockers. Installing the package alone is not live recovery proof.