📝 Refresh Almanac runtime and release docs

almanac/architecture/automation/contributor-automation.md

@@ -48,6 +48,6 @@ That remediation flow has its own validation boundary. If it applies any fixes,
 
 ## Agent Guards
 
-`dangerfile.ts` protects `README.md`, `LICENSE`, `CLAUDE.md`, `assets/**`, `.claude/**`, `.codex/**`, and `.cursor/**` from agent-authored pull request changes [@dangerfile]. It detects agent authorship from the PR author, commit author names and emails, and `Co-authored-by` trailers [@dangerfile]. The Danger workflow runs `npm run danger:ci` on pull requests to `main` after installing dependencies on Node 24 [@danger-workflow] [@package].
+`dangerfile.ts` protects `README.md`, `CONTRIBUTING.md`, `CODE_OF_CONDUCT.md`, `LICENSE`, `CLAUDE.md`, `docs/**`, `assets/**`, `.claude/**`, `.codex/**`, and `.cursor/**` from agent-authored pull request changes [@dangerfile]. It detects agent authorship from the PR author, commit author names and emails, and `Co-authored-by` trailers [@dangerfile]. The Danger workflow runs `npm run danger:ci` on pull requests to `main` after installing dependencies on Node 24 [@danger-workflow] [@package].
 
 For detected-agent pull requests, Danger fails any changed path matching `PROTECTED_PATHS` and warns when the policy files `dangerfile.ts` or `.github/workflows/danger.yml` change [@dangerfile]. It also enforces inline `agent-guard:off` and `agent-guard:on` regions: invalid marker nesting fails, agent-authored changes inside protected regions fail, and marker changes warn for explicit review [@dangerfile]. The lookup version of this contract belongs in [Protected Paths](../../reference/contributor/protected-paths).

almanac/architecture/release/release-automation.md

@@ -18,23 +18,23 @@ Rudder's release automation is split between a publishing workflow on `main` and
 
 ## Version As Release Input
 
-`package.json` is the only version input for the workflows. The package is currently `@ruddercode/rudder-plugin` at version `0.1.0`, and both workflows turn that version into tag `rudder-plugin-v0.1.0` at runtime [@package-json] [@publish-workflow] [@release-alert]. The release-alert workflow performs the same package-name, version, tag, npm, and GitHub Release checks on pull requests to `main` [@release-alert].
+`package.json` is the only version input for the workflows. Both workflows read the manifest version at runtime and turn it into tag `rudder-plugin-v<version>` [@package-json] [@publish-workflow] [@release-alert]. The release-alert workflow performs the same package-name, version, tag, npm, and GitHub Release checks on pull requests to `main` [@release-alert].
 
 Both workflows reject any package name other than `@ruddercode/rudder-plugin` before attempting release work [@publish-workflow] [@release-alert]. This is a release invariant: changing the package name affects plugin marketplace metadata, npm publication, and tag naming, so evaluate it through [Prepare Package Release](../../guides/release/prepare-package-release) and [Rudder Plugin Package](../tooling/plugin-package).
 
 ## Publish Job
 
-The publish workflow runs on pushes to `main` and on manual dispatch, with a single `publish-rudder-plugin` concurrency group and `contents: write` plus `id-token: write` permissions [@publish-workflow]. Its first shell step checks whether npmjs.org already has `name@version`, whether the package exists at all, whether the `rudder-plugin-v<version>` tag exists, and whether a GitHub Release exists for that tag [@publish-workflow].
+The publish workflow runs on pushes to `main` and on manual dispatch, with a single `publish-rudder-plugin` concurrency group and `contents: write` plus `id-token: write` permissions [@publish-workflow]. Its first shell step checks whether npmjs.org already has `name@version`, whether the `rudder-plugin-v<version>` tag exists, and whether a GitHub Release exists for that tag [@publish-workflow].
 
-Those checks create separate flags for each artifact. A missing npmjs.org version enables npm publishing, a missing package plus missing version marks a bootstrap publish, a missing tag enables tag creation, and a missing GitHub Release enables release creation [@publish-workflow]. When any artifact is missing, the job upgrades npm, installs dependencies with `npm ci`, validates the plugin package, publishes to npmjs.org when needed, pushes the plugin tag when needed, and creates the GitHub Release with generated notes when needed [@publish-workflow].
+Those checks create separate flags for each artifact. A missing npmjs.org version enables npm publishing, a missing tag enables tag creation, and a missing GitHub Release enables release creation [@publish-workflow]. When any artifact is missing, the job upgrades npm, installs dependencies with `npm ci`, writes release telemetry defaults into `src/telemetry-build-config.ts`, validates the plugin package with `DO_NOT_TRACK=1`, publishes to npmjs.org when needed with Trusted Publishing, pushes the plugin tag when needed, and creates the GitHub Release with generated notes when needed [@publish-workflow].
 
-The package scripts are part of that path because the publish workflow validates with `npm run check:agent-layout`, `npm run typecheck`, `npm test`, `npm run build`, and `npm pack --dry-run`, while `npm publish` also uses the package lifecycle in `package.json` [@publish-workflow] [@package-json]. The exact scripts are listed in [Package Scripts](../../reference/tooling/package-scripts).
+The package scripts are part of that path because the publish workflow validates with `npm run check:agent-layout`, `npm run typecheck`, `npm test`, `npm run build`, and `npm pack --dry-run`, while `npm publish` also uses the package lifecycle in `package.json` [@publish-workflow] [@package-json]. The GitHub Release title is `Rudder v<version>` even though the tag remains `rudder-plugin-v<version>` [@publish-workflow]. The exact scripts are listed in [Package Scripts](../../reference/tooling/package-scripts).
 
 ## Release Alert Job
 
 The release-alert workflow runs on pull requests opened, synchronized, or reopened against `main`; it has read permissions for contents plus write permission for pull request comments [@release-alert]. It uses a per-PR `plugin-release-alert-<pr-number>` concurrency group and cancels in-progress runs on newer pushes, so the latest branch state owns the PR warning [@release-alert].
 
-The alert job mirrors the publish checks without publishing. It checks npmjs.org, bootstrap-token need, tag creation, and GitHub Release creation, then sets `will_release=true` when any release artifact would be created after merge [@release-alert]. A GitHub Script step searches for an existing comment containing `<!-- release-alert -->` and updates it, or creates one when missing [@release-alert]. The comment warns when merge would publish the plugin, use the bootstrap token, create the plugin tag, or create a release, and it switches to a no-release message when all plugin artifacts already exist for the version [@release-alert].
+The alert job mirrors the publish checks without publishing. It checks npmjs.org, tag creation, and GitHub Release creation, then sets `will_release=true` when any release artifact would be created after merge [@release-alert]. A GitHub Script step searches for an existing comment containing `<!-- release-alert -->` and updates it, or creates one when missing [@release-alert]. The comment warns when merge would publish the plugin, create the plugin tag, or create a release, and it switches to a no-release message when all plugin artifacts already exist for the version [@release-alert].
 
 ## Release Model
 

almanac/architecture/runtime/local-state.md

@@ -1,27 +1,30 @@
 ---
 title: "Local State"
-summary: "Rudder keeps runtime state in a user-scoped home directory that owns the SQLite database, prompt-capture preference, telemetry identity file, and dashboard port defaults."
+summary: "Rudder keeps runtime state in a user-scoped home directory that owns the SQLite database, telemetry identity, update-check cache, backups, and dashboard port defaults."
 topics: [architecture, runtime, local-state, sqlite, prompt-capture]
 sources:
   - id: db-client
     type: file
     path: src/db/client.ts
-  - id: prompt-control
-    type: file
-    path: src/prompt-control.ts
   - id: telemetry
     type: file
     path: src/telemetry.ts
+  - id: update-script
+    type: file
+    path: skills/rudder/scripts/update.mjs
+  - id: backup-script
+    type: file
+    path: skills/rudder/scripts/backup-tests.mjs
   - id: gitignore
     type: file
     path: .gitignore
 ---
 
-Rudder local state is the small persistent runtime surface that exists outside the source tree. The database client resolves a Rudder home directory from `RUDDER_HOME` or `~/.rudder`, stores the SQLite database at `rudder.db`, opens that database through a process-wide singleton, enables WAL journaling, applies generated Drizzle migrations, and then exposes Drizzle access [@db-client]. The same home directory stores the prompt-capture disable marker and anonymous telemetry identity file, so [Prompt Branch Store](prompt-branch-store), prompt controls, and [Telemetry](telemetry) share one local state root [@prompt-control] [@telemetry]. Runtime artifacts live under the selected Rudder home instead of a repository-local state directory, while the source tree keeps only the code and configuration that derive those paths [@db-client] [@gitignore].
+Rudder local state is the small persistent runtime surface that exists outside the source tree. The database client resolves a Rudder home directory from `RUDDER_HOME` or `~/.rudder`, stores the SQLite database at `rudder.db`, opens that database through a process-wide singleton, enables WAL journaling, applies generated Drizzle migrations, and then exposes Drizzle access [@db-client]. The same home directory stores the anonymous telemetry identity file, the update-check cache, and skill-created test-reset backups, so [Prompt Branch Store](prompt-branch-store), [Telemetry](telemetry), and the skill runtime share one local state root [@telemetry] [@update-script] [@backup-script]. Runtime artifacts live under the selected Rudder home instead of a repository-local state directory, while the source tree keeps only the code and configuration that derive those paths [@db-client] [@gitignore].
 
 ## State Root
 
-`rudderHome()` is the owner of the local state path. It returns `process.env.RUDDER_HOME` when that value is present, and otherwise joins the operating-system home directory with `.rudder` [@db-client]. `dbPath()` derives the database location by joining that root with `rudder.db`, and `promptCaptureDisabledPath()` derives the persistent capture preference as `<rudderHome()>/prompt-capture-disabled` [@db-client] [@prompt-control].
+`rudderHome()` is the owner of the local state path. It returns `process.env.RUDDER_HOME` when that value is present, and otherwise joins the operating-system home directory with `.rudder` [@db-client]. `dbPath()` derives the database location by joining that root with `rudder.db` [@db-client]. The update helper derives `<rudderHome()>/update-state.json`, and the backup helper writes test-reset backups under `<rudderHome()>/backups/` [@update-script] [@backup-script].
 
 The source tree does not carry a repo-local state directory convention. The repository ignore file covers dependencies, build output, generated backups, logs, environment files, coverage, and editor files, but it does not define a `.rudder/` workspace cache [@gitignore]. The runtime code instead creates the selected Rudder home directory directly, so changing the state location is an environment-variable choice rather than a working-tree layout change [@db-client].
 
@@ -37,4 +40,4 @@ Migration application is deliberately part of the open flow. `openDb()` derives
 
 ## Shared Boundary
 
-Local state currently covers the SQLite database path, the prompt-capture preference marker, the telemetry identity file, and the dashboard port default. Telemetry builds `identity.json` under `rudderHome()`, reads an existing `{ id }` value when present, and writes a generated UUID there on a best-effort basis when it needs a new anonymous installation identity [@telemetry]. The important invariant is that runtime code should derive persistent paths from `rudderHome()` instead of inventing new repository-local locations. That keeps [Telemetry](telemetry), [Prompt Branch Store](prompt-branch-store), prompt controls, and the environment-variable reference aligned around the same state root [@db-client] [@prompt-control].
+Local state currently covers the SQLite database path, the telemetry identity file, the update-check cache, skill backup directories, and the dashboard port default. Telemetry builds `identity.json` under `rudderHome()`, reads an existing `{ id }` value when present, and writes a generated UUID there on a best-effort basis when it needs a new anonymous installation identity [@telemetry]. The update helper writes `update-state.json` atomically through a temporary file with mode `0600` after creating the Rudder home with mode `0700` [@update-script]. The important invariant is that runtime code should derive persistent paths from `rudderHome()` instead of inventing new repository-local locations. That keeps [Telemetry](telemetry), [Prompt Branch Store](prompt-branch-store), the update helper, and the environment-variable reference aligned around the same state root [@db-client] [@telemetry] [@update-script].