chore[notask]: prefix custom Cursor skills with qv-

.cursor/rules/devops/agentic-automation.mdc

@@ -21,9 +21,9 @@ Rules governing agent-driven automation in DevOps surfaces — Cursor skills, ho
 ## Read-only default
 
 - **Skills default to read-only.** Mutations to the local repo, the remote (GitHub state, releases, branch protection), or external systems (cloud, k8s, Slack, Asana) require explicit user opt-in per call.
-- **The `pr-review` pattern is the reference**: print the exact command, wait for user confirmation, then execute. Never auto-execute.
+- **The `qv-pr-review` pattern is the reference**: print the exact command, wait for user confirmation, then execute. Never auto-execute.
 - **No silent file mutations** to the user's working tree. If a skill writes outside `/tmp/` or its own `_lib/` cache, it announces what and why first.
-- **No silent git mutations**: forbidden without explicit user instruction — `git switch`, `git checkout` (any ref/file), `git reset`, `git restore`, `git stash`, `git pull`, `git merge`, `git rebase`, `git cherry-pick`, `git clean`, `gh pr checkout`. (See `.cursor/skills/pr-review/SKILL.md` "Safety rules" section for the canonical wording.)
+- **No silent git mutations**: forbidden without explicit user instruction — `git switch`, `git checkout` (any ref/file), `git reset`, `git restore`, `git stash`, `git pull`, `git merge`, `git rebase`, `git cherry-pick`, `git clean`, `gh pr checkout`. (See `.cursor/skills/qv-pr-review/SKILL.md` "Safety rules" section for the canonical wording.)
 
 ## Plan-then-apply for state changes
 
@@ -44,7 +44,7 @@ Rules governing agent-driven automation in DevOps surfaces — Cursor skills, ho
 ## Idempotency
 
 - **Skills must be idempotent.** Re-running with identical inputs produces the same effect, or reports "already applied" without making changes.
-- **Generated artifacts (NOTICE files, changelog entries, generated docs) sort deterministically** so re-runs produce byte-identical output and clean diffs (the `notice-generate` skill is the reference).
+- **Generated artifacts (NOTICE files, changelog entries, generated docs) sort deterministically** so re-runs produce byte-identical output and clean diffs (the `qv-notice-generate` skill is the reference).
 - **Network-dependent skills** cache responses where reasonable (`/tmp/<skill>-<id>.json`) and document staleness.
 
 ## Validation before success
@@ -78,15 +78,15 @@ Rules governing agent-driven automation in DevOps surfaces — Cursor skills, ho
 
 ## Bounded resource use
 
-- **Bound shell-call count per skill.** State an explicit budget in the skill's "Efficiency rules" section (the `pr-review` skill caps at ~5–8 calls; treat that as the ceiling unless justified).
+- **Bound shell-call count per skill.** State an explicit budget in the skill's "Efficiency rules" section (the `qv-pr-review` skill caps at ~5–8 calls; treat that as the ceiling unless justified).
 - **Cache fetched data once per session** (`/tmp/<skill>-<id>.json`). Never re-fetch the same PR/run/file in one session.
 - **Use dedicated tools, not shell.** `Read` instead of `cat`/`head`/`tail`, `Grep` instead of `grep`/`rg`, `Glob` instead of `find`, `Write`/`StrReplace` instead of `echo >` / heredoc.
 - **Prefer scripts in `_lib/<domain>/`** over inline shell pipelines. Scripts are testable, deterministic, and share-able across skills.
 
 ## Skill authoring conventions
 
 - **One skill = one capability.** Don't compose unrelated workflows in a single skill.
-- **Naming**: kebab-case, prefix-grouped (`devops-*`, `sdk-*`, `addon-*`, `pr-*`).
+- **Naming**: kebab-case, all custom repo skills use the `qv-` prefix, then pod or domain segment (`qv-devops-*`, `qv-sdk-*`, `qv-addon-*`, `qv-pr-*`).
 - **`SKILL.md` body ≤ 500 lines.** Push detail to `references/*.md` and link one level deep.
 - **Slash-command discoverability**: include `/<skill-name>` in the description.
 - **Auto-invoke vs explicit**: `disable-model-invocation: true` for any skill that mutates state, posts publicly, or runs slow / expensive tooling. Auto-invoke only for read-only, fast, contextually obvious helpers.

.cursor/rules/devops/main.mdc

@@ -29,7 +29,7 @@ Owns CI/CD pipelines (GitHub Actions workflows + composite actions), repo-wide a
 | `.cursor/rules/devops/secrets-and-credentials.mdc` | `.github/**`, `scripts/**` | Anything that can read, derive, or transmit a secret |
 | `.cursor/rules/devops/agentic-automation.mdc` | `.cursor/skills/**`, `.cursor/rules/**`, `.cursor/hooks/**`, `.github/workflows/**`, `.github/scripts/**`, `scripts/**` | Authoring skills, hooks, MCP integrations, or any AI-driven DevOps workflow |
 
-Commit message and PR title format guidance lives in the [`devops-pr-create`](../../skills/devops-pr-create/SKILL.md) skill — invoke it explicitly rather than autoloading a rule for it (skill-only on purpose, to keep the format spec out of the always-on context window).
+Commit message and PR title format guidance lives in the [`qv-devops-pr-create`](../../skills/qv-devops-pr-create/SKILL.md) skill — invoke it explicitly rather than autoloading a rule for it (skill-only on purpose, to keep the format spec out of the always-on context window).
 
 ## Quick rules — surface-level reminders
 
@@ -82,4 +82,4 @@ These are the bar; the deep-dive lives in the topic files above.
 
 ## Commit messages and PR titles
 
-The full format spec, allowed prefixes, allowed tags, validation regex, and quality guidelines live in the [`devops-pr-create`](../../skills/devops-pr-create/SKILL.md) skill. Invoke `/devops-pr-create` rather than relying on this rule — kept out of the auto-load context window on purpose.
+The full format spec, allowed prefixes, allowed tags, validation regex, and quality guidelines live in the [`qv-devops-pr-create`](../../skills/qv-devops-pr-create/SKILL.md) skill. Invoke `/qv-devops-pr-create` rather than relying on this rule — kept out of the auto-load context window on purpose.

.cursor/rules/registry-server/main.mdc

@@ -73,4 +73,4 @@ Use `lib/logger.js`. Avoid logging sensitive information (keys, tokens).
 
 ### Autobase + HyperDB Multi-Writer Patterns
 
-When working with Autobase, HyperDB, multi-writer patterns, or peer discovery, use the `registry-autobase-patterns` skill.
+When working with Autobase, HyperDB, multi-writer patterns, or peer discovery, use the `qv-registry-autobase-patterns` skill.

.cursor/rules/sdk/docs-freshness.mdc

@@ -37,7 +37,7 @@ This rule monitors two packages that have dedicated knowledge-base documentation
 
 | Source area | Documentation |
 |---|---|
-| `lib/registry-service.js`, `scripts/build-db-spec.js`, `shared/spec/` | `.cursor/skills/registry-autobase-patterns/` |
+| `lib/registry-service.js`, `scripts/build-db-spec.js`, `shared/spec/` | `.cursor/skills/qv-registry-autobase-patterns/` |
 
 ## What to do
 

.cursor/rules/sdk/sdk-pod-packages.mdc

@@ -30,8 +30,10 @@ globs:
 
 ## Skill Naming Convention
 
-Skills for SDK pod should use the `sdk-` prefix:
+Skills for SDK pod should use the `qv-sdk-` prefix:
 
-- `sdk-pr-create`
-- `sdk-changelog`
-- `sdk-release`
+- `qv-sdk-pr-create`
+- `qv-sdk-changelog`
+- `qv-sdk-backmerge`
+- `qv-sdk-e2e-create`
+- `qv-sdk-pr-status`

.cursor/rules/sdk/skill-authoring-guidelines.mdc

@@ -2,7 +2,7 @@
 description: Guidelines for creating SDK pod Cursor skills
 alwaysApply: false
 globs:
-  - .cursor/skills/sdk-**
+  - .cursor/skills/qv-sdk-**
 ---
 
 # Skill Authoring Guidelines
@@ -25,10 +25,12 @@ Guidelines for creating effective Cursor skills, based on the [Agent Skills spec
 
 ## Naming Convention
 
-**Use pod prefix for skill directories:**
+**Use `qv-` prefix plus pod segment for skill directories:**
 
-- SDK pod skills: `sdk-pr-create`, `sdk-changelog`, `sdk-release`
-- Other pods: `infra-deploy`, `platform-migration`
+- SDK pod skills: `qv-sdk-pr-create`, `qv-sdk-changelog`, `qv-sdk-backmerge`
+- DevOps pod: `qv-devops-pr-create`, `qv-devops-pr-review`
+- Addon pod: `qv-addon-changelog`, `qv-addon-pr-create`
+- Cross-pod PR workflow: `qv-pr-review`, `qv-pr-test`, `qv-pr-mine`
 
 **If pod is unclear, ask the user.**
 
@@ -42,7 +44,7 @@ Guidelines for creating effective Cursor skills, based on the [Agent Skills spec
 
 **Prefer gerund form or action-oriented names:**
 
-- Good: `processing-pdfs`, `analyzing-data`, `sdk-pr-create`
+- Good: `processing-pdfs`, `analyzing-data`, `qv-sdk-pr-create`
 - Avoid: `helper`, `utils`, `tools`
 
 ## Required Frontmatter
@@ -328,7 +330,7 @@ Choose one term and use it throughout:
 
 ### Vague Skill Names
 
-- Good: `sdk-pr-create`, `processing-pdfs`
+- Good: `qv-sdk-pr-create`, `processing-pdfs`
 - Bad: `helper`, `utils`, `stuff`
 
 ## Degrees of Freedom

.cursor/skills/_lib/pr-skills/README.md

@@ -6,9 +6,9 @@ Cross-cutting daily workflow helpers live next door in
 
 This directory does not contain a `SKILL.md`; it is not a Cursor skill itself. The user-facing skills live under `.cursor/skills/`:
 
-- `pr-review/` — generic single-PR review. Pod-agnostic.
-- `pr-test/` — generic single-PR local validation. Pod-agnostic.
-- `pr-mine/` — cross-pod "my open PRs" dashboard. Discovers every pod under `.github/teams/` and routes per-PR ping logic to the owning pod.
+- `qv-pr-review/` — generic single-PR review. Pod-agnostic.
+- `qv-pr-test/` — generic single-PR local validation. Pod-agnostic.
+- `qv-pr-mine/` — cross-pod "my open PRs" dashboard. Discovers every pod under `.github/teams/` and routes per-PR ping logic to the owning pod.
 - `<pod>-pr-status/` — per-pod team dashboard, thin wrapper invoking the shared script with `--pod <name> --mode team`.
 
 ## Files
@@ -17,19 +17,19 @@ This directory does not contain a `SKILL.md`; it is not a Cursor skill itself. T
 - [`pr-activity.mjs`](pr-activity.mjs) — shared PR data collector/classifier used by `pr-status.mjs` and daily workflow tooling. It fetches open PRs, resolves pod ownership, computes review state, stale/re-review/ready groups, and merge-conflict flags.
 - [`team.mjs`](team.mjs) — team-metadata loader. `loadTeam(pod)` reads a single pod; `discoverPods()` enumerates every `.github/teams/<pod>.json`; `findPodForFiles(files, pods)` returns the pod that owns a PR's touched files (first match wins).
 - [`slack.mjs`](slack.mjs) — Slack-handle map loader. File lives at `~/.config/qvac-pr-skills/slack.json`, schema `{ map, pendingReview }`. Bootstraps missing entries from `gh api users/<login>` and parks newly seeded logins on `pendingReview` so the skill workflow can confirm them with the user.
-- [`worktree.mjs`](worktree.mjs) — worktree manager for `/pr-review` and `/pr-test`. `resolvePR` (gh-resolved baseRefName, fail-fast — does NOT default to main), `fetchPRRefs` (single fetch for both PR head and base ref), `ensureWorktreeSynced` (sync mode, in-place `reset --hard`; preserves untracked artifacts at the same SHA and runs `clean -fdx` only after SHA drift), `lockPR` (per-PR flock), `computePatch` (3-dot diff to `/tmp/pr-<num>.patch`), `cleanupCache` (LRU keep 3).
-- [`worktree-prepare.mjs`](worktree-prepare.mjs) — CLI entry for `/pr-review`'s worktree mode. Success: prints `WORKTREE_PATH`, `HEAD_SHA`, `PATCH_PATH`, `BASE_REF` on stdout. Any failure: prints `WORKTREE_FALLBACK=<reason>` on stderr and exits 0 so the agent transparently falls back to API-only mode.
-- [`pr-test-discover.mjs`](pr-test-discover.mjs) — discovery CLI for `/pr-test`. Orchestrates PR metadata/patch loading, package discovery, and JSON manifest output. It never uses `git diff`/`git status` against the worktree for classification.
+- [`worktree.mjs`](worktree.mjs) — worktree manager for `/qv-pr-review` and `/qv-pr-test`. `resolvePR` (gh-resolved baseRefName, fail-fast — does NOT default to main), `fetchPRRefs` (single fetch for both PR head and base ref), `ensureWorktreeSynced` (sync mode, in-place `reset --hard`; preserves untracked artifacts at the same SHA and runs `clean -fdx` only after SHA drift), `lockPR` (per-PR flock), `computePatch` (3-dot diff to `/tmp/pr-<num>.patch`), `cleanupCache` (LRU keep 3).
+- [`worktree-prepare.mjs`](worktree-prepare.mjs) — CLI entry for `/qv-pr-review`'s worktree mode. Success: prints `WORKTREE_PATH`, `HEAD_SHA`, `PATCH_PATH`, `BASE_REF` on stdout. Any failure: prints `WORKTREE_FALLBACK=<reason>` on stderr and exits 0 so the agent transparently falls back to API-only mode.
+- [`pr-test-discover.mjs`](pr-test-discover.mjs) — discovery CLI for `/qv-pr-test`. Orchestrates PR metadata/patch loading, package discovery, and JSON manifest output. It never uses `git diff`/`git status` against the worktree for classification.
 - [`pr-test-generic.mjs`](pr-test-generic.mjs) — package-agnostic discovery helpers: PR JSON/patch loading, patch-status parsing, package path detection, package.json script discovery, generic package classification, and non-SDK recommendation logic.
-- [`pr-test-sdk.mjs`](pr-test-sdk.mjs) — SDK-specific `/pr-test` heuristics: SDK pod package detection, SDK example command generation, related SDK examples, related `tests-qvac` filters, and SDK e2e setup recommendations.
+- [`pr-test-sdk.mjs`](pr-test-sdk.mjs) — SDK-specific `/qv-pr-test` heuristics: SDK pod package detection, SDK example command generation, related SDK examples, related `tests-qvac` filters, and SDK e2e setup recommendations.
 
 ## Modes
 
 | Mode | Pod scope | What it shows | Used by |
 |---|---|---|---|
 | `team` | required (`--pod`) | All open PRs touching the pod's `ownedPaths` that still need reviews. Three sections: needs-your-re-review, stale (>3d), needs-review. PRs with `mergeable: CONFLICTING` are flagged with `⚠️ MERGE CONFLICTS!`. Pass `--authors pod` to additionally scope the dashboard to PRs authored by pod-roster members; non-roster authors touching pod paths are surfaced in a separate "Excluded" section. | `<pod>-pr-status` |
 | `review` | required (`--pod`) | The current user's personal review queue: PRs needing their first review, plus PRs where their review was dismissed. | (currently unused; available for a future skill) |
-| `my` | optional (`--pod`); cross-pod by default | The current user's own open PRs grouped by merge readiness. Per-PR pod resolution drives ping logic. Emits copy-paste Slack ping messages for missing reviewers. | `pr-mine` |
+| `my` | optional (`--pod`); cross-pod by default | The current user's own open PRs grouped by merge readiness. Per-PR pod resolution drives ping logic. Emits copy-paste Slack ping messages for missing reviewers. | `qv-pr-mine` |
 
 ## CLI
 
@@ -75,11 +75,11 @@ from the local `upstream` remote.
 
    `ownedPaths` are prefix-matched against changed-file paths to decide whether a PR is "owned" by this pod. Use trailing slashes.
 
-2. Create the per-pod dashboard skill by copying `.cursor/skills/sdk-pr-status/` to `.cursor/skills/<pod>-pr-status/`. Inside the copy, update the SKILL.md frontmatter (`name:`, `description:`) and the script invocation in the `## Usage` block to swap `--pod sdk` for `--pod <pod>`. No other changes required.
+2. Create the per-pod dashboard skill by copying `.cursor/skills/qv-sdk-pr-status/` to `.cursor/skills/qv-<pod>-pr-status/`. Inside the copy, update the SKILL.md frontmatter (`name:`, `description:`) and the script invocation in the `## Usage` block to swap `--pod sdk` for `--pod <pod>`. No other changes required.
 
-3. `pr-mine`, `pr-review`, and `pr-test` are NOT pod-specific and do not need duplication. They live at `.cursor/skills/pr-mine/`, `.cursor/skills/pr-review/`, and `.cursor/skills/pr-test/` and discover the new pod automatically.
+3. `qv-pr-mine`, `qv-pr-review`, and `qv-pr-test` are NOT pod-specific and do not need duplication. They live at `.cursor/skills/qv-pr-mine/`, `.cursor/skills/qv-pr-review/`, and `.cursor/skills/qv-pr-test/` and discover the new pod automatically.
 
-4. The first time anyone on the new pod runs `/pr-mine`, the shared script auto-fills `~/.config/qvac-pr-skills/slack.json` with `gh api users/<login>` names for the newly added logins and emits `SLACK_VALIDATION_REQUIRED <N>` on stderr, prompting the skill workflow to drive a confirm-or-correct flow with the user.
+4. The first time anyone on the new pod runs `/qv-pr-mine`, the shared script auto-fills `~/.config/qvac-pr-skills/slack.json` with `gh api users/<login>` names for the newly added logins and emits `SLACK_VALIDATION_REQUIRED <N>` on stderr, prompting the skill workflow to drive a confirm-or-correct flow with the user.
 
 ## Slack-handle map (per-user, never committed)
 
@@ -91,7 +91,7 @@ from the local `upstream` remote.
     "pendingReview": ["<github-login>"]
   }
   ```
-- The script appends to `pendingReview` whenever it auto-fills a new entry. The `pr-mine` SKILL workflow consumes the pending list, presents each entry to the user via `AskQuestion`, applies corrections, and clears `pendingReview` once validation is done.
+- The script appends to `pendingReview` whenever it auto-fills a new entry. The `qv-pr-mine` SKILL workflow consumes the pending list, presents each entry to the user via `AskQuestion`, applies corrections, and clears `pendingReview` once validation is done.
 - Edit the file directly at any time — the script never overwrites entries already in `map`, only adds new ones.
 
 ## Worktree cache (per-user, never committed)
@@ -100,7 +100,7 @@ from the local `upstream` remote.
 - One subdirectory per PR num: `pr-<num>/` (sync mode — single path per PR, kept in sync via `fetch + reset --hard` when SHA advances).
 - One lock file per PR num: `pr-<num>.lock`. Held by `worktree-prepare.mjs` for the duration of fetch + sync; released before printing the path. Same-PR concurrent invocations serialize on this lock.
 - LRU eviction: only the 3 most recently touched worktrees are retained. Older entries are removed via `git worktree remove --force` + `rm -rf`.
-- Used by `/pr-review` (read-only source inspection) and `/pr-test` (package manager/build/test commands inside the isolated cache).
+- Used by `/qv-pr-review` (read-only source inspection) and `/qv-pr-test` (package manager/build/test commands inside the isolated cache).
 - Dirty handling:
   - Same SHA: `reset --hard HEAD` resets tracked-file edits and preserves untracked build/test artifacts (`node_modules`, `dist`, native `build/`, logs).
   - SHA drift: `reset --hard refs/pr/<num>/head` followed by `clean -fdx` so artifacts from the old PR head are not reused against the new commit.

.cursor/skills/_lib/pr-skills/pr-test-discover.mjs

@@ -1,6 +1,6 @@
 #!/usr/bin/env node
 //
-// Discovery helper for /pr-test.
+// Discovery helper for /qv-pr-test.
 //
 // Usage:
 //   node pr-test-discover.mjs <PR-URL> --worktree <WORKTREE_PATH> \
@@ -9,7 +9,7 @@
 // The file list and per-file status come from committed PR state only:
 // `/tmp/pr-<num>.json` (`gh pr view --json files`) plus the patch emitted by
 // worktree-prepare.mjs. This helper never uses git diff/status against the
-// worktree, which may contain untracked build artifacts from /pr-test.
+// worktree, which may contain untracked build artifacts from /qv-pr-test.
 
 import { readFileSync } from "node:fs";
 import { dirname, join, relative } from "node:path";

.cursor/skills/_lib/pr-skills/team.mjs

@@ -64,7 +64,7 @@ export function loadTeam(pod) {
 }
 
 // Discover every pod registered under .github/teams/. Used by cross-pod
-// modes (e.g. /pr-mine) where the user's PRs may span multiple pods.
+// modes (e.g. /qv-pr-mine) where the user's PRs may span multiple pods.
 // Returns an array of team objects (same shape as loadTeam(pod)).
 export function discoverPods() {
   const dir = teamsDir();

.cursor/skills/_lib/pr-skills/worktree-prepare.mjs

@@ -1,6 +1,6 @@
 #!/usr/bin/env node
 //
-// CLI entry for the default-on worktree mode of /pr-review.
+// CLI entry for the default-on worktree mode of /qv-pr-review.
 //
 // Usage:
 //   node worktree-prepare.mjs <PR-URL>