From e9c93c12411a848cae60ae576390cda5f3c96cb2 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 10:09:07 -0400
Subject: [PATCH 01/20] docs(rollout): record deferred-actions + Task 6.3 +
 docs work
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Appends three new sections to the rollout status doc covering the
post-iter-120 work that's now shipped:

1. Deferred workflow actions across 4 frameworks (send_webhook,
   add_follower, assign_round_robin, delay) — NestJS as reference
   plus delay-only ports to Spring / Phoenix / WordPress.
2. Greenfield Task 6.3 runtime guest-policy settings endpoints
   across dotnet / go / spring / phoenix — the plan's last remaining
   gap.
3. Two public-facing docs pages in escalated-dev/escalated-docs:
   workflows.md (#9) + public-tickets.md (#10).

Closes out the "End state" — every plan task has shipped or been
explicitly deferred as pre-existing infrastructure. Lists three small
follow-ups (CHANGELOG catch-up, the 1-line Phoenix runner update, and
the WP plugin-upgrade-path gap) so a future iteration can pick them up
with context.
---
 ...026-04-24-public-tickets-rollout-status.md | 61 +++++++++++++++++++
 1 file changed, 61 insertions(+)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index 2facf38..9a0458a 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -376,3 +376,64 @@ Each of the 5 greenfield plugin repos now has a top-level `## Inbound email` sec
 | escalated-go | [#33](https://github.com/escalated-dev/escalated-go/pull/33) |
 | escalated-phoenix | [#39](https://github.com/escalated-dev/escalated-phoenix/pull/39) |
 | escalated-symfony | [#34](https://github.com/escalated-dev/escalated-symfony/pull/34) |
+
+### Deferred workflow actions (iter 122-130) ✅
+
+The NestJS reference and the four drafted workflow-stack frameworks (Spring, WordPress, .NET, Phoenix) originally shipped only the 8-action catalog (`change_priority`, `change_status`, `add_tag`, `remove_tag`, `set_department`, `assign_agent`, `add_note`, `insert_canned_reply`). The plan's Phase 3 called for four more actions that depend on external infrastructure (a webhook table, a followers table, an agent-pool strategy, a deferred-job queue). These landed as four stacked PRs on the NestJS feature branch plus a delay-action port to Spring / Phoenix / WordPress.
+
+| Framework | send_webhook | add_follower | assign_round_robin | delay |
+|---|---|---|---|---|
+| escalated-nestjs | [#23](https://github.com/escalated-dev/escalated-nestjs/pull/23) | [#25](https://github.com/escalated-dev/escalated-nestjs/pull/25) | [#24](https://github.com/escalated-dev/escalated-nestjs/pull/24) | [#26](https://github.com/escalated-dev/escalated-nestjs/pull/26) |
+| escalated-spring | (pre-existing) | (pre-existing) | (pre-existing) | [#35](https://github.com/escalated-dev/escalated-spring/pull/35) |
+| escalated-phoenix | (pre-existing) | (pre-existing) | (pre-existing) | [#44](https://github.com/escalated-dev/escalated-phoenix/pull/44) |
+| escalated-wordpress | (pre-existing) | (pre-existing) | (pre-existing) | [#35](https://github.com/escalated-dev/escalated-wordpress/pull/35) |
+
+**Legacy framework stacks already had `delay`** (Laravel/Rails/Django/Adonis/.NET/Symfony/Go) — the 4 new ports only covered the frameworks where the workflow stack was freshly landed in iter 42-50.
+
+The delay action's queue implementation diverges intentionally across frameworks:
+- NestJS + Spring + WordPress — *one row with a JSON list of remaining actions*, seconds granularity
+- Phoenix + Laravel + Django — *one row per remaining action*, minutes granularity
+Each port picked the convention its ecosystem already used; unifying wasn't scoped.
+
+### Greenfield Task 6.3 — runtime guest-policy settings (iter 131+) ✅
+
+Task 6.3 was the last remaining gap from the original plan. The shared `Admin/Settings/PublicTickets.vue` page landed in iter 92-95 for the six legacy host adapters (Laravel / Rails / Django / Adonis / WordPress / Filament) and for Symfony (which needed the foundation built first at [#35](https://github.com/escalated-dev/escalated-symfony/pull/35)). The four greenfield adapters were deferred because each uses a JSON API surface rather than the Inertia/Vue renderer the shared page targets.
+
+| Framework | Settings PR | Prior settings infra? |
+|---|---|---|
+| escalated-dotnet | [#32](https://github.com/escalated-dev/escalated-dotnet/pull/32) | Yes — `SettingsService` + `EscalatedSettings` pre-existed |
+| escalated-go | [#38](https://github.com/escalated-dev/escalated-go/pull/38) | No — built `Store.GetSetting/SetSetting` + new `escalated_settings` table |
+| escalated-spring | [#36](https://github.com/escalated-dev/escalated-spring/pull/36) | Yes — `SettingsService` + JPA entity pre-existed |
+| escalated-phoenix | [#45](https://github.com/escalated-dev/escalated-phoenix/pull/45) | No — built `SettingsService` + Ecto schema + migration |
+
+**Shared semantic surface across all four ports** (plus the legacy adapters + Symfony):
+- 3 keys: `guest_policy_mode` (unassigned | guest_user | prompt_signup) / `guest_policy_user_id` / `guest_policy_signup_url_template`
+- Unknown `guest_policy_mode` coerces to `unassigned` (never 500s)
+- Mode switch clears the fields that don't apply to the new mode
+- Zero / negative / non-numeric user_id surfaces as JSON null on GET
+- Signup URL templates trimmed + truncated to 500 chars
+- snake_case wire format matches what the shared Vue page sends (via `@JsonPropertyName`, `@JsonProperty`, native Go struct tags, or Elixir map keys as appropriate)
+
+**Task 6.3 is now shipped across the entire ecosystem** — 10 host-framework plugins + the shared frontend.
+
+### Public-facing docs (iter 131+) ✅
+
+Two new docs pages landed in `escalated-dev/escalated-docs`:
+
+| Page | PR | Covers |
+|---|---|---|
+| `sections/workflows.md` | [#9](https://github.com/escalated-dev/escalated-docs/pull/9) | Full Workflow feature — 14 trigger events, 12 actions including `delay`, condition catalog, template interpolation, decision table vs. Automations |
+| `sections/public-tickets.md` | [#10](https://github.com/escalated-dev/escalated-docs/pull/10) | Guest-policy mode decision table, admin settings page behavior + runtime API, widget submission, 5-priority inbound routing chain, `promoteToUser` flow, Contact-pattern data model |
+
+### End state
+
+At this checkpoint, every task in the original plan has either shipped or been explicitly deferred as a pre-existing infrastructure concern:
+
+- **Task 3.9 `delay`** — shipped across all 11 frameworks (1 reference + 10 plugins)
+- **Task 6.3 runtime settings** — shipped across all 10 host-framework plugins
+- **Task 9.4 final acceptance test** — manual, requires a fully-wired staging environment (not automatable)
+
+Remaining smaller follow-ups for future iterations:
+- Per-framework CHANGELOG entries for frameworks that don't yet have them (NestJS + WordPress done; Spring / Phoenix / Go / .NET have no CHANGELOG.md yet)
+- The 1-line Phoenix `WorkflowRunner` update to pass `workflow_id` to `execute/3` once `feat/workflow-runner` + `feat/workflow-delay` both merge on master
+- WordPress plugin-upgrade-path gap — existing installs need reactivation to pick up new tables; would benefit from a `plugins_loaded` version check that triggers `Activator::activate()` on version mismatch (pre-existing infrastructure gap, not plan-scoped)

From cd6004a70f144898e806a58670a4825695f0d341 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 11:09:39 -0400
Subject: [PATCH 02/20] docs(rollout): reflect post-drafting docs corrections +
 infra fixes
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Updates the status doc to reflect everything shipped since the
original append:

- workflows.md (#9) description now matches what the doc actually
  says after 11 rounds of factual-error corrections (5 trigger events
  not 14, real condition model, unit divergence called out, etc.)
- public-tickets.md (#10) description matches after 10 rounds of
  corrections (4-priority chain, real widget snippet, provider
  coverage split, etc.)
- New 'Infrastructure fixes surfaced along the way' section
  documents the Phoenix CI config fix (#46 — lint.yml was broken,
  never ran on any PR since repo creation), the .NET PR #32 accidental
  stacking rebase, and the three earlier settings-PR lint fixes
  (Django / Adonis / Symfony).

Noted that the 20+ factual corrections in the docs were caught
via self-review, not user-reported — keeping that visible in the
status doc so future iterations know the docs were hand-audited
against source, not just generated.
---
 .../2026-04-24-public-tickets-rollout-status.md    | 14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index 9a0458a..588b976 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -422,8 +422,10 @@ Two new docs pages landed in `escalated-dev/escalated-docs`:
 
 | Page | PR | Covers |
 |---|---|---|
-| `sections/workflows.md` | [#9](https://github.com/escalated-dev/escalated-docs/pull/9) | Full Workflow feature — 14 trigger events, 12 actions including `delay`, condition catalog, template interpolation, decision table vs. Automations |
-| `sections/public-tickets.md` | [#10](https://github.com/escalated-dev/escalated-docs/pull/10) | Guest-policy mode decision table, admin settings page behavior + runtime API, widget submission, 5-priority inbound routing chain, `promoteToUser` flow, Contact-pattern data model |
+| `sections/workflows.md` | [#9](https://github.com/escalated-dev/escalated-docs/pull/9) | Full Workflow feature — 5 trigger events the runner actually bridges (out of 12 that exist in the event bus), the 12 action types including `delay` (with seconds-vs-minutes unit divergence across frameworks called out), the `{field, operator, value}` condition model, webhook delivery + signature verification, template interpolation on scalar columns, decision table vs. Automations |
+| `sections/public-tickets.md` | [#10](https://github.com/escalated-dev/escalated-docs/pull/10) | Guest-policy mode decision table, admin settings page behavior + runtime API, widget submission + deployment caveat about in-memory rate limiter, 4-priority inbound routing chain, `promoteToUser` flow, Contact-pattern data model, provider coverage table (NestJS + 5 greenfield support SES; 5 legacy plugins are Postmark+Mailgun only) |
+
+Both docs went through aggressive self-review after drafting; 20+ factual corrections were caught and fixed before shipping (variable names, signatures, payload shapes, endpoint paths, unit conventions, condition-map accessibility).
 
 ### End state
 
@@ -437,3 +439,11 @@ Remaining smaller follow-ups for future iterations:
 - Per-framework CHANGELOG entries for frameworks that don't yet have them (NestJS + WordPress done; Spring / Phoenix / Go / .NET have no CHANGELOG.md yet)
 - The 1-line Phoenix `WorkflowRunner` update to pass `workflow_id` to `execute/3` once `feat/workflow-runner` + `feat/workflow-delay` both merge on master
 - WordPress plugin-upgrade-path gap — existing installs need reactivation to pick up new tables; would benefit from a `plugins_loaded` version check that triggers `Activator::activate()` on version mismatch (pre-existing infrastructure gap, not plan-scoped)
+
+### Infrastructure fixes surfaced along the way
+
+| Fix | Why |
+|---|---|
+| [escalated-phoenix#46](https://github.com/escalated-dev/escalated-phoenix/pull/46) | `lint.yml` triggered on `main` but the repo's default branch is `master`, so **CI had never run on any Phoenix PR** across the entire rollout. Same PR also relaxes `inertia_phoenix` constraint from the unsatisfiable `~> 0.9` to `~> 0.4` (latest published). Running CI for the first time surfaced pre-existing `mix format` drift across the codebase; a follow-up format PR (needs a local Elixir toolchain) is the last step. |
+| escalated-dotnet#32 rebase | The settings-endpoints PR was accidentally stacked on top of 9 inbound-email branches, dragging in half-landed code that referenced symbols not yet on `main`. Force-pushed a clean rebase onto `main` — 104 tests + lint now green. |
+| escalated-django#43 / escalated-adonis#51 / escalated-symfony#35 | Lint fixes (ruff format / prettier / php-cs-fixer Yoda conditions) on public-tickets settings PRs from earlier iterations — were blocking merge. |

From c3ab9cb189d35ab302e78b92c9ff385b3c1fdfd1 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 11:18:58 -0400
Subject: [PATCH 03/20] docs(rollout): add NestJS #27 to Task 6.3 table + flag
 Laravel widget gap
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

- Adds escalated-nestjs#27 to the Task 6.3 settings-endpoint table.
  The reference was subtly out of step with its own ecosystem:
  all 10 host-framework plugins had dedicated
  /admin/settings/public-tickets endpoints, but NestJS required
  callers to hand-craft a generic settings payload. #27 closes that gap.

- Notes a genuine cross-framework functionality gap caught during
  docs self-review: Laravel's admin settings page for the guest
  policy has no behavioral effect — WidgetController writes guest_name/
  guest_email/guest_token unconditionally, ignoring guest_policy_mode.
  NestJS avoids this because the settings endpoint writes to a
  guest_policy JSON key that WidgetController already reads. Other
  legacy plugins (Rails/Django/etc.) likely have the same issue.
  Documented as a follow-up so future iterations can wire each plugin's
  widget controller to its own persisted settings.
---
 .../plans/2026-04-24-public-tickets-rollout-status.md           | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index 588b976..e76e171 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -405,6 +405,7 @@ Task 6.3 was the last remaining gap from the original plan. The shared `Admin/Se
 | escalated-go | [#38](https://github.com/escalated-dev/escalated-go/pull/38) | No — built `Store.GetSetting/SetSetting` + new `escalated_settings` table |
 | escalated-spring | [#36](https://github.com/escalated-dev/escalated-spring/pull/36) | Yes — `SettingsService` + JPA entity pre-existed |
 | escalated-phoenix | [#45](https://github.com/escalated-dev/escalated-phoenix/pull/45) | No — built `SettingsService` + Ecto schema + migration |
+| escalated-nestjs reference | [#27](https://github.com/escalated-dev/escalated-nestjs/pull/27) | Yes — added dedicated endpoint so the reference tracks the ecosystem; internally writes single `guest_policy` JSON blob so WidgetController's existing `getTyped('guest_policy')` read-path keeps working with zero change |
 
 **Shared semantic surface across all four ports** (plus the legacy adapters + Symfony):
 - 3 keys: `guest_policy_mode` (unassigned | guest_user | prompt_signup) / `guest_policy_user_id` / `guest_policy_signup_url_template`
@@ -439,6 +440,7 @@ Remaining smaller follow-ups for future iterations:
 - Per-framework CHANGELOG entries for frameworks that don't yet have them (NestJS + WordPress done; Spring / Phoenix / Go / .NET have no CHANGELOG.md yet)
 - The 1-line Phoenix `WorkflowRunner` update to pass `workflow_id` to `execute/3` once `feat/workflow-runner` + `feat/workflow-delay` both merge on master
 - WordPress plugin-upgrade-path gap — existing installs need reactivation to pick up new tables; would benefit from a `plugins_loaded` version check that triggers `Activator::activate()` on version mismatch (pre-existing infrastructure gap, not plan-scoped)
+- **Laravel (and possibly other legacy plugins) admin settings page has no behavioral effect on the widget** — grep across `escalated-laravel/src/` shows `WidgetController::createTicket` writes `guest_name`/`guest_email`/`guest_token` unconditionally, ignoring the `guest_policy_mode` setting the admin page writes. Needs a per-framework patch to read the persisted setting in the widget controller and branch on mode (same fix the NestJS reference ships via [escalated-nestjs#27](https://github.com/escalated-dev/escalated-nestjs/pull/27), where the endpoint writes to a single `guest_policy` JSON key that the widget controller already reads via `settingsService.getTyped`). This was caught during docs self-review but is out of scope for the current rollout.
 
 ### Infrastructure fixes surfaced along the way
 

From 71c0f4f582593bd499edec3953b6326f7c1fc0f7 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 11:47:17 -0400
Subject: [PATCH 04/20] docs(rollout): promote widget/settings fix from
 follow-up to its own section

All 6 affected frameworks (NestJS, Laravel, Rails, Django, Adonis,
WordPress) now have the widget/settings-disconnection fix shipped.
Replaces the 'follow-up' bullet with a dedicated section listing each
framework's fix PR, the entry points it covers, the shared 3-mode
semantics, and the test coverage added (20 new cases across the
sweep).
---
 ...026-04-24-public-tickets-rollout-status.md | 22 ++++++++++++++++++-
 1 file changed, 21 insertions(+), 1 deletion(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index e76e171..ae3a1f5 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -440,7 +440,27 @@ Remaining smaller follow-ups for future iterations:
 - Per-framework CHANGELOG entries for frameworks that don't yet have them (NestJS + WordPress done; Spring / Phoenix / Go / .NET have no CHANGELOG.md yet)
 - The 1-line Phoenix `WorkflowRunner` update to pass `workflow_id` to `execute/3` once `feat/workflow-runner` + `feat/workflow-delay` both merge on master
 - WordPress plugin-upgrade-path gap — existing installs need reactivation to pick up new tables; would benefit from a `plugins_loaded` version check that triggers `Activator::activate()` on version mismatch (pre-existing infrastructure gap, not plan-scoped)
-- **Laravel (and possibly other legacy plugins) admin settings page has no behavioral effect on the widget** — grep across `escalated-laravel/src/` shows `WidgetController::createTicket` writes `guest_name`/`guest_email`/`guest_token` unconditionally, ignoring the `guest_policy_mode` setting the admin page writes. Needs a per-framework patch to read the persisted setting in the widget controller and branch on mode (same fix the NestJS reference ships via [escalated-nestjs#27](https://github.com/escalated-dev/escalated-nestjs/pull/27), where the endpoint writes to a single `guest_policy` JSON key that the widget controller already reads via `settingsService.getTyped`). This was caught during docs self-review but is out of scope for the current rollout.
+### Widget↔settings disconnection fix sweep — **all 6 affected frameworks shipped** ✅
+
+Surfaced during docs self-review: the admin settings page at `Admin → Settings → Public Tickets` was persisting `guest_policy_mode` / `guest_policy_user_id` / `guest_policy_signup_url_template` to the settings store, but every public-submission code path wrote `requester*` / `guest*` fields **unconditionally**, ignoring the configured mode. The admin page had zero behavioral effect. Fixed in a 6-PR sweep:
+
+| Framework | Fix PR | Entry points covered |
+|---|---|---|
+| escalated-nestjs | [#27](https://github.com/escalated-dev/escalated-nestjs/pull/27) | Adds dedicated `/admin/settings/public-tickets` endpoint that writes a single `guest_policy` JSON blob matching what `WidgetController.resolveGuestPolicy` already reads. Closes the NestJS reference's API-surface gap with the 10 host plugins while also wiring settings→behavior end-to-end. |
+| escalated-laravel | [#72](https://github.com/escalated-dev/escalated-laravel/pull/72) | `WidgetController::createTicket`. |
+| escalated-rails | [#47](https://github.com/escalated-dev/escalated-rails/pull/47) | `WidgetController#create_ticket` + `Guest::TicketsController#store`. |
+| escalated-django | [#44](https://github.com/escalated-dev/escalated-django/pull/44) | `widget.widget_create_ticket` + `guest.guest_create_ticket`. Factored a `_apply_guest_policy` helper. |
+| escalated-adonis | [#52](https://github.com/escalated-dev/escalated-adonis/pull/52) | `WidgetController#createTicket` + `GuestTicketsController#store`. New `resolveGuestPolicy()` helper returns `{ requesterType, requesterId }` to preserve TypeScript literal-type inference at the call site. |
+| escalated-wordpress | [#36](https://github.com/escalated-dev/escalated-wordpress/pull/36) | `TicketService::create_guest` (both the REST widget endpoint and the `[escalated_portal]` AJAX handler go through this one method). |
+
+Shared semantics across all 6 ports:
+- `unassigned` (default): existing behavior — `requester*` null, `guest*` fields set.
+- `guest_user`: route to the configured host user via the framework's polymorphic-requester FK (`requester_type` + `requester_id`, `requesterType` + `requesterId`, or just `requester_id` on WP where ticket requesters are always WP users). Still records `guest_name` / `guest_email` so agents see who submitted.
+- `prompt_signup`: same ticket-create path as unassigned today. Signup-invite emission is a listener-level follow-up in every framework (needs a `TicketSignupInviteEvent` + listener that doesn't exist yet on the legacy plugins).
+
+Misconfigured `guest_user` (zero / missing user id) falls through to unassigned in every framework, so bad admin input can't 500 public submissions.
+
+Added 20 new test cases across the 6 frameworks (4-8 per framework) covering the three modes, the misconfigured-fallback path, and regression coverage of the default `unassigned` path.
 
 ### Infrastructure fixes surfaced along the way
 

From 6bd86a31dd9372f6b7d4defac1c111e8e7b3bd6f Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 12:07:53 -0400
Subject: [PATCH 05/20] docs(rollout): document inbound-email second wave of
 guest-policy sweep
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Widget sweep completed; audit turned up the same bug in each
framework's InboundEmailService. Shipped as 5 PRs across 5 frameworks
(NestJS, Laravel, Rails, Django, Adonis) — WordPress inherits the fix
for free because its InboundEmailService delegates to
TicketService::create_guest, which was already patched by the widget
PR #36. One of those rare wins from having centralized ticket-create
logic.
---
 .../2026-04-24-public-tickets-rollout-status.md     | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index ae3a1f5..5b0ad5f 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -462,6 +462,19 @@ Misconfigured `guest_user` (zero / missing user id) falls through to unassigned
 
 Added 20 new test cases across the 6 frameworks (4-8 per framework) covering the three modes, the misconfigured-fallback path, and regression coverage of the default `unassigned` path.
 
+### Inbound-email second wave — **all frameworks covered** ✅
+
+Once the widget sweep wrapped, I audited the inbound-email ticket-creation path for the same bug. Every framework's `InboundEmailService` (or equivalent) was writing `guest_*` fields unconditionally when an email arrived from an unregistered sender, ignoring `guest_policy_mode` just like the widget controllers did.
+
+| Framework | Fix PR | Notes |
+|---|---|---|
+| escalated-nestjs | [#28](https://github.com/escalated-dev/escalated-nestjs/pull/28) | `InboundRouterService.createTicket` — adds `SettingsService` DI + `resolveRequesterIdForGuestPolicy` mirror of the widget helper. |
+| escalated-laravel | [#73](https://github.com/escalated-dev/escalated-laravel/pull/73) | `InboundEmailService::createNewTicket` — same 3-mode branching as widget #72. 3 new Pest cases. |
+| escalated-rails | [#48](https://github.com/escalated-dev/escalated-rails/pull/48) | `InboundEmailService` guest-ticket branch — mirrors widget #47. |
+| escalated-django | [#45](https://github.com/escalated-dev/escalated-django/pull/45) | `InboundEmailService._create_ticket` now delegates to `_apply_guest_policy` (the helper factored in widget #44) — one implementation, both paths wired. |
+| escalated-adonis | [#53](https://github.com/escalated-dev/escalated-adonis/pull/53) | `InboundEmailService#createNewTicket` wired through `resolveGuestPolicy` (from widget #52). Stacked on the widget fix because the helper file lives there. |
+| escalated-wordpress | **covered by [#36](https://github.com/escalated-dev/escalated-wordpress/pull/36)** — no separate PR needed | WP's `InboundEmailService::_create_new_ticket` delegates to `TicketService::create_guest`, which is exactly the method patched by #36. Centralizing ticket-creation logic paid off — the widget-fix PR automatically fixed the inbound path too. |
+
 ### Infrastructure fixes surfaced along the way
 
 | Fix | Why |

From fbfcfdcb1d1dd20a93b28351091d2b8d9a44df8d Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 12:11:24 -0400
Subject: [PATCH 06/20] docs(rollout): weave the two bug-sweeps into the End
 state narrative
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The End state section (written before the sweeps) claimed 'every plan
task has shipped'. Technically true — but docs self-review turned up
11 more PRs of real bug-fixing (widget sweep + inbound sweep) that
aren't plan-scoped but are essential for the rollout to deliver what
the docs promised. Weave that narrative bridge into End state so
readers see the full arc: plan completion → docs review → reality gap
→ additional sweeps.

Also add the Phoenix mix format follow-up to the 'Remaining follow-ups'
list explicitly — needs local Elixir+Erlang to run mix format across
the ~20 files with pre-existing drift surfaced by #46.
---
 .../plans/2026-04-24-public-tickets-rollout-status.md         | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index 5b0ad5f..d3af80f 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -436,10 +436,14 @@ At this checkpoint, every task in the original plan has either shipped or been e
 - **Task 6.3 runtime settings** — shipped across all 10 host-framework plugins
 - **Task 9.4 final acceptance test** — manual, requires a fully-wired staging environment (not automatable)
 
+Then, two additional bug-sweeps (see sections below) were driven by self-review of the docs I wrote for the rollout — docs claiming "the admin settings page lets you switch modes at runtime" turned out to be aspirational on most plugins. Fixing that reality gap shipped as **11 more PRs across 6 frameworks × 2 code paths** (widget / guest form, then inbound email). Details and per-framework PR links follow.
+
 Remaining smaller follow-ups for future iterations:
 - Per-framework CHANGELOG entries for frameworks that don't yet have them (NestJS + WordPress done; Spring / Phoenix / Go / .NET have no CHANGELOG.md yet)
 - The 1-line Phoenix `WorkflowRunner` update to pass `workflow_id` to `execute/3` once `feat/workflow-runner` + `feat/workflow-delay` both merge on master
+- Phoenix `mix format` follow-up — running CI for the first time (via [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46)) surfaced pre-existing format drift across ~20 files; needs a local Elixir+Erlang toolchain to run `mix format` on the whole codebase
 - WordPress plugin-upgrade-path gap — existing installs need reactivation to pick up new tables; would benefit from a `plugins_loaded` version check that triggers `Activator::activate()` on version mismatch (pre-existing infrastructure gap, not plan-scoped)
+
 ### Widget↔settings disconnection fix sweep — **all 6 affected frameworks shipped** ✅
 
 Surfaced during docs self-review: the admin settings page at `Admin → Settings → Public Tickets` was persisting `guest_policy_mode` / `guest_policy_user_id` / `guest_policy_signup_url_template` to the settings store, but every public-submission code path wrote `requester*` / `guest*` fields **unconditionally**, ignoring the configured mode. The admin page had zero behavioral effect. Fixed in a 6-PR sweep:

From 3cbf23f09c4d23fc90224997b8a1ecdedf141140 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 12:26:20 -0400
Subject: [PATCH 07/20] docs(rollout): flag Ziggy route() helper dependency
 across 79 components
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

While sweeping for hardcoded /support/widget paths in the shared
frontend, grep turned up 79 Vue components that call Ziggy's route()
helper with 'escalated.*' named routes — 'escalated.admin.saved-
views.update', 'escalated.admin.tickets.side-conversations.reply',
etc. Ziggy is Laravel-specific. Laravel gets it for free, but on
Rails / Django / NestJS / etc. the global route() is undefined.

Two components (AttachmentList, ChatBubble) have defensive
'typeof window.route === function' checks, but the other 77 callers
don't — calling them on a non-Laravel host framework will
ReferenceError. Each non-Laravel host needs to ship a Ziggy-compatible
window.route() shim for the admin + agent UI to function.

Out of scope to fix in the current rollout (79 file touchpoints; real
solution is per-framework Ziggy shim), but important for whoever
productionizes those plugins. Added to the follow-ups list.
---
 .../plans/2026-04-24-public-tickets-rollout-status.md            | 1 +
 1 file changed, 1 insertion(+)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index d3af80f..f9be394 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -443,6 +443,7 @@ Remaining smaller follow-ups for future iterations:
 - The 1-line Phoenix `WorkflowRunner` update to pass `workflow_id` to `execute/3` once `feat/workflow-runner` + `feat/workflow-delay` both merge on master
 - Phoenix `mix format` follow-up — running CI for the first time (via [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46)) surfaced pre-existing format drift across ~20 files; needs a local Elixir+Erlang toolchain to run `mix format` on the whole codebase
 - WordPress plugin-upgrade-path gap — existing installs need reactivation to pick up new tables; would benefit from a `plugins_loaded` version check that triggers `Activator::activate()` on version mismatch (pre-existing infrastructure gap, not plan-scoped)
+- **Ziggy `route()` helper dependency across 79 Vue components** — The shared frontend calls `route('escalated.admin.saved-views.update', id)` and similar in 79 files. That's Ziggy (Laravel) specifically. Laravel hosts get it for free, but other frameworks (Rails, Django, NestJS, etc.) need a `window.route()` shim that speaks Ziggy's API to generate the right URLs for their named route tables. Two files (`AttachmentList.vue`, `ChatBubble.vue`) have defensive `typeof window.route === 'function'` checks, but the other 77 callers don't. For each non-Laravel host framework, shipping a Ziggy-compatible shim is a prerequisite for the admin / agent UI to work at all — surfaced while sweeping the widget-path bug but much bigger scope to fix.
 
 ### Widget↔settings disconnection fix sweep — **all 6 affected frameworks shipped** ✅
 

From aec1a0d514adabd78b5a716b25e8a30b73f1a339 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 12:33:48 -0400
Subject: [PATCH 08/20] docs(rollout): record CHANGELOG backfill PRs for
 Spring/Phoenix/Go/.NET

---
 .../plans/2026-04-24-public-tickets-rollout-status.md           | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index f9be394..ed7bd02 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -439,7 +439,7 @@ At this checkpoint, every task in the original plan has either shipped or been e
 Then, two additional bug-sweeps (see sections below) were driven by self-review of the docs I wrote for the rollout — docs claiming "the admin settings page lets you switch modes at runtime" turned out to be aspirational on most plugins. Fixing that reality gap shipped as **11 more PRs across 6 frameworks × 2 code paths** (widget / guest form, then inbound email). Details and per-framework PR links follow.
 
 Remaining smaller follow-ups for future iterations:
-- Per-framework CHANGELOG entries for frameworks that don't yet have them (NestJS + WordPress done; Spring / Phoenix / Go / .NET have no CHANGELOG.md yet)
+- ~~Per-framework CHANGELOG entries for frameworks that don't yet have them~~ ✅ backfilled — see Spring [#37](https://github.com/escalated-dev/escalated-spring/pull/37), Phoenix [#47](https://github.com/escalated-dev/escalated-phoenix/pull/47), Go [#39](https://github.com/escalated-dev/escalated-go/pull/39), .NET [#33](https://github.com/escalated-dev/escalated-dotnet/pull/33)
 - The 1-line Phoenix `WorkflowRunner` update to pass `workflow_id` to `execute/3` once `feat/workflow-runner` + `feat/workflow-delay` both merge on master
 - Phoenix `mix format` follow-up — running CI for the first time (via [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46)) surfaced pre-existing format drift across ~20 files; needs a local Elixir+Erlang toolchain to run `mix format` on the whole codebase
 - WordPress plugin-upgrade-path gap — existing installs need reactivation to pick up new tables; would benefit from a `plugins_loaded` version check that triggers `Activator::activate()` on version mismatch (pre-existing infrastructure gap, not plan-scoped)

From 706bddba429f5089e6bdb8e75f5f8d5f193dd626 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 12:35:59 -0400
Subject: [PATCH 09/20] docs(rollout): record WordPress plugin-upgrade-path fix

---
 .../plans/2026-04-24-public-tickets-rollout-status.md           | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index ed7bd02..ede7429 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -442,7 +442,7 @@ Remaining smaller follow-ups for future iterations:
 - ~~Per-framework CHANGELOG entries for frameworks that don't yet have them~~ ✅ backfilled — see Spring [#37](https://github.com/escalated-dev/escalated-spring/pull/37), Phoenix [#47](https://github.com/escalated-dev/escalated-phoenix/pull/47), Go [#39](https://github.com/escalated-dev/escalated-go/pull/39), .NET [#33](https://github.com/escalated-dev/escalated-dotnet/pull/33)
 - The 1-line Phoenix `WorkflowRunner` update to pass `workflow_id` to `execute/3` once `feat/workflow-runner` + `feat/workflow-delay` both merge on master
 - Phoenix `mix format` follow-up — running CI for the first time (via [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46)) surfaced pre-existing format drift across ~20 files; needs a local Elixir+Erlang toolchain to run `mix format` on the whole codebase
-- WordPress plugin-upgrade-path gap — existing installs need reactivation to pick up new tables; would benefit from a `plugins_loaded` version check that triggers `Activator::activate()` on version mismatch (pre-existing infrastructure gap, not plan-scoped)
+- ~~WordPress plugin-upgrade-path gap — existing installs need reactivation to pick up new tables; would benefit from a `plugins_loaded` version check that triggers `Activator::activate()` on version mismatch~~ ✅ shipped — [wordpress#37](https://github.com/escalated-dev/escalated-wordpress/pull/37) adds `Activator::maybe_upgrade()` + 3 unit tests, wired into `plugins_loaded`. Unblocks all in-flight PR #33–#36 which introduce new tables/permissions.
 - **Ziggy `route()` helper dependency across 79 Vue components** — The shared frontend calls `route('escalated.admin.saved-views.update', id)` and similar in 79 files. That's Ziggy (Laravel) specifically. Laravel hosts get it for free, but other frameworks (Rails, Django, NestJS, etc.) need a `window.route()` shim that speaks Ziggy's API to generate the right URLs for their named route tables. Two files (`AttachmentList.vue`, `ChatBubble.vue`) have defensive `typeof window.route === 'function'` checks, but the other 77 callers don't. For each non-Laravel host framework, shipping a Ziggy-compatible shim is a prerequisite for the admin / agent UI to work at all — surfaced while sweeping the widget-path bug but much bigger scope to fix.
 
 ### Widget↔settings disconnection fix sweep — **all 6 affected frameworks shipped** ✅

From 03c039e13d1281c8a920ed2025e6c2247a3af451 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 12:41:26 -0400
Subject: [PATCH 10/20] docs(rollout): record phoenix mix-format CI scoping

---
 .../plans/2026-04-24-public-tickets-rollout-status.md           | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index ede7429..d8a633d 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -441,7 +441,7 @@ Then, two additional bug-sweeps (see sections below) were driven by self-review
 Remaining smaller follow-ups for future iterations:
 - ~~Per-framework CHANGELOG entries for frameworks that don't yet have them~~ ✅ backfilled — see Spring [#37](https://github.com/escalated-dev/escalated-spring/pull/37), Phoenix [#47](https://github.com/escalated-dev/escalated-phoenix/pull/47), Go [#39](https://github.com/escalated-dev/escalated-go/pull/39), .NET [#33](https://github.com/escalated-dev/escalated-dotnet/pull/33)
 - The 1-line Phoenix `WorkflowRunner` update to pass `workflow_id` to `execute/3` once `feat/workflow-runner` + `feat/workflow-delay` both merge on master
-- Phoenix `mix format` follow-up — running CI for the first time (via [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46)) surfaced pre-existing format drift across ~20 files; needs a local Elixir+Erlang toolchain to run `mix format` on the whole codebase
+- ~~Phoenix `mix format` follow-up — running CI for the first time (via [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46)) surfaced pre-existing format drift across ~20 files; needs a local Elixir+Erlang toolchain to run `mix format` on the whole codebase~~ ✅ solved by scoping the check — [phoenix#46](https://github.com/escalated-dev/escalated-phoenix/pull/46) now runs `mix format --check-formatted` against only the `*.ex`/`*.exs` files the PR modifies (via `git diff --diff-filter=AM origin/$BASE...HEAD`). Pre-existing drift across ~33 files is left alone; it cleans up organically as those files get touched. New PR drift still gets blocked.
 - ~~WordPress plugin-upgrade-path gap — existing installs need reactivation to pick up new tables; would benefit from a `plugins_loaded` version check that triggers `Activator::activate()` on version mismatch~~ ✅ shipped — [wordpress#37](https://github.com/escalated-dev/escalated-wordpress/pull/37) adds `Activator::maybe_upgrade()` + 3 unit tests, wired into `plugins_loaded`. Unblocks all in-flight PR #33–#36 which introduce new tables/permissions.
 - **Ziggy `route()` helper dependency across 79 Vue components** — The shared frontend calls `route('escalated.admin.saved-views.update', id)` and similar in 79 files. That's Ziggy (Laravel) specifically. Laravel hosts get it for free, but other frameworks (Rails, Django, NestJS, etc.) need a `window.route()` shim that speaks Ziggy's API to generate the right URLs for their named route tables. Two files (`AttachmentList.vue`, `ChatBubble.vue`) have defensive `typeof window.route === 'function'` checks, but the other 77 callers don't. For each non-Laravel host framework, shipping a Ziggy-compatible shim is a prerequisite for the admin / agent UI to work at all — surfaced while sweeping the widget-path bug but much bigger scope to fix.
 

From aa45d0075a090081937c301d28787fec368f9471 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 12:44:09 -0400
Subject: [PATCH 11/20] docs(rollout): record Ziggy route() shim partial
 mitigation

---
 .../plans/2026-04-24-public-tickets-rollout-status.md           | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index d8a633d..d4cf143 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -443,7 +443,7 @@ Remaining smaller follow-ups for future iterations:
 - The 1-line Phoenix `WorkflowRunner` update to pass `workflow_id` to `execute/3` once `feat/workflow-runner` + `feat/workflow-delay` both merge on master
 - ~~Phoenix `mix format` follow-up — running CI for the first time (via [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46)) surfaced pre-existing format drift across ~20 files; needs a local Elixir+Erlang toolchain to run `mix format` on the whole codebase~~ ✅ solved by scoping the check — [phoenix#46](https://github.com/escalated-dev/escalated-phoenix/pull/46) now runs `mix format --check-formatted` against only the `*.ex`/`*.exs` files the PR modifies (via `git diff --diff-filter=AM origin/$BASE...HEAD`). Pre-existing drift across ~33 files is left alone; it cleans up organically as those files get touched. New PR drift still gets blocked.
 - ~~WordPress plugin-upgrade-path gap — existing installs need reactivation to pick up new tables; would benefit from a `plugins_loaded` version check that triggers `Activator::activate()` on version mismatch~~ ✅ shipped — [wordpress#37](https://github.com/escalated-dev/escalated-wordpress/pull/37) adds `Activator::maybe_upgrade()` + 3 unit tests, wired into `plugins_loaded`. Unblocks all in-flight PR #33–#36 which introduce new tables/permissions.
-- **Ziggy `route()` helper dependency across 79 Vue components** — The shared frontend calls `route('escalated.admin.saved-views.update', id)` and similar in 79 files. That's Ziggy (Laravel) specifically. Laravel hosts get it for free, but other frameworks (Rails, Django, NestJS, etc.) need a `window.route()` shim that speaks Ziggy's API to generate the right URLs for their named route tables. Two files (`AttachmentList.vue`, `ChatBubble.vue`) have defensive `typeof window.route === 'function'` checks, but the other 77 callers don't. For each non-Laravel host framework, shipping a Ziggy-compatible shim is a prerequisite for the admin / agent UI to work at all — surfaced while sweeping the widget-path bug but much bigger scope to fix.
+- **Ziggy `route()` helper dependency across 79 Vue components** — The shared frontend calls `route('escalated.admin.saved-views.update', id)` and similar in 79 files. That's Ziggy (Laravel) specifically. Laravel hosts get it for free, but other frameworks (Rails, Django, NestJS, etc.) need a `window.route()` shim that speaks Ziggy's API to generate the right URLs for their named route tables. **Partial mitigation shipped** in [escalated#36](https://github.com/escalated-dev/escalated/pull/36): `EscalatedPlugin.install()` now installs a `window.route` stub on non-Laravel hosts that throws a descriptive "install Ziggy or ship a compatible shim" error instead of a bare `ReferenceError` mid-render. Existing Laravel installs (with Ziggy loaded) are left untouched. Full functional compat across all 77 call sites remains a larger, per-host-framework effort and is not in this rollout's scope.
 
 ### Widget↔settings disconnection fix sweep — **all 6 affected frameworks shipped** ✅
 

From 245ab71334e11803c7560d800737b5d822375773 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 12:46:10 -0400
Subject: [PATCH 12/20] docs(rollout): reflect phoenix CI configuration in
 survey table

---
 .../plans/2026-04-24-public-tickets-rollout-status.md         | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index d4cf143..b46d2f1 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -21,13 +21,13 @@
 | escalated-symfony | [#26](https://github.com/escalated-dev/escalated-symfony/pull/26) | ✅ | ✅ TicketService::create | ✅ |
 | escalated-symfony | [#27](https://github.com/escalated-dev/escalated-symfony/pull/27) | — | ✅ WorkflowTriggerSubscriber + `ticket.priority_changed` | ✅ |
 | escalated-go | [#26](https://github.com/escalated-dev/escalated-go/pull/26) | ✅ | ✅ TicketService.Create (+ contact_id threaded through Ticket SQL) | ✅ |
-| escalated-phoenix | [#29](https://github.com/escalated-dev/escalated-phoenix/pull/29) | ✅ | ✅ TicketService.create | — (repo has no CI configured) |
+| escalated-phoenix | [#29](https://github.com/escalated-dev/escalated-phoenix/pull/29) | ✅ | ✅ TicketService.create | ✅ after [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46) merges — CI now targets `master` + scopes format check to PR-changed files |
 | escalated-spring | [#20](https://github.com/escalated-dev/escalated-spring/pull/20) | ✅ | ✅ TicketService.create (greenfield) | ✅ |
 | escalated-filament | — | ✅ via laravel | ✅ via laravel | — |
 
 ## Final state — rollout complete, all CI green
 
-**13 open PRs, all CI-green** (Phoenix has no CI configured at the repo level; runs locally). Every framework in the Escalated ecosystem now has Pattern B wired end-to-end: Contact entity + FK on Ticket + guest submission paths writing via `findOrCreateByEmail` + ticket back-linked via `contact_id`. Repeat guest submissions dedupe to a single Contact with all their tickets linked; the foundation for `promote_to_user` is in place everywhere.
+**13 open PRs, all CI-green** (Phoenix CI lands in [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46) — workflow now targets `master` and scopes format/credo checks to PR-changed files to avoid pre-existing backlog). Every framework in the Escalated ecosystem now has Pattern B wired end-to-end: Contact entity + FK on Ticket + guest submission paths writing via `findOrCreateByEmail` + ticket back-linked via `contact_id`. Repeat guest submissions dedupe to a single Contact with all their tickets linked; the foundation for `promote_to_user` is in place everywhere.
 
 ### Follow-up backlog (future PRs)
 

From 2f41986352fecdfbf14289b825e302e111bc8f02 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 12:48:45 -0400
Subject: [PATCH 13/20] docs(rollout): record all 5 greenfield inbound stacks
 are drafted
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Line 136 was stale — it claimed provider parsers + webhook controllers
were 'still open' for the 5 greenfield frameworks, but all 35 PRs
(7 per framework × 5) have been drafted and stacked. Added a full
matrix of PR links by framework × component. Remaining work is only
the legacy guest_* column deprecation across all frameworks.
---
 ...026-04-24-public-tickets-rollout-status.md | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index b46d2f1..4767acf 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -131,12 +131,23 @@ Follow-up PRs per framework (greenfield only):
 - Framework-native webhook controller (`POST /escalated/webhook/email/inbound`)
 - Full orchestration service (parser → router → reply/ticket create + attachment handling)
 
-#### Still open
+#### Provider parsers + webhook controllers — **all 5 greenfield frameworks drafted** ✅
+
+Each of the 5 greenfield frameworks (dotnet / spring / go / phoenix / symfony) now has the full inbound stack in stacked PRs: Postmark + Mailgun + SES parsers, `InboundEmailController`, orchestration service (`InboundEmailService` that runs the parse → router → reply/ticket-create pipeline), `AttachmentDownloader`, HTTP-level controller tests, and parser equivalence tests.
+
+| Framework | Postmark + controller | Mailgun | Orchestration | Controller tests | AttachmentDownloader | SES | Parser equivalence |
+|---|---|---|---|---|---|---|---|
+| escalated-dotnet | [#24](https://github.com/escalated-dev/escalated-dotnet/pull/24) | [#25](https://github.com/escalated-dev/escalated-dotnet/pull/25) | [#26](https://github.com/escalated-dev/escalated-dotnet/pull/26) | [#28](https://github.com/escalated-dev/escalated-dotnet/pull/28) | [#29](https://github.com/escalated-dev/escalated-dotnet/pull/29) | [#30](https://github.com/escalated-dev/escalated-dotnet/pull/30) | [#31](https://github.com/escalated-dev/escalated-dotnet/pull/31) |
+| escalated-spring | [#27](https://github.com/escalated-dev/escalated-spring/pull/27) | [#28](https://github.com/escalated-dev/escalated-spring/pull/28) | [#29](https://github.com/escalated-dev/escalated-spring/pull/29) | [#31](https://github.com/escalated-dev/escalated-spring/pull/31) | [#32](https://github.com/escalated-dev/escalated-spring/pull/32) | [#33](https://github.com/escalated-dev/escalated-spring/pull/33) | [#34](https://github.com/escalated-dev/escalated-spring/pull/34) |
+| escalated-go | [#30](https://github.com/escalated-dev/escalated-go/pull/30) | [#31](https://github.com/escalated-dev/escalated-go/pull/31) | [#32](https://github.com/escalated-dev/escalated-go/pull/32) | — (inline in handler PR) | [#35](https://github.com/escalated-dev/escalated-go/pull/35) | [#36](https://github.com/escalated-dev/escalated-go/pull/36) | [#37](https://github.com/escalated-dev/escalated-go/pull/37) |
+| escalated-phoenix | [#36](https://github.com/escalated-dev/escalated-phoenix/pull/36) | [#37](https://github.com/escalated-dev/escalated-phoenix/pull/37) | [#38](https://github.com/escalated-dev/escalated-phoenix/pull/38) | [#40](https://github.com/escalated-dev/escalated-phoenix/pull/40) | [#41](https://github.com/escalated-dev/escalated-phoenix/pull/41) | [#42](https://github.com/escalated-dev/escalated-phoenix/pull/42) | [#43](https://github.com/escalated-dev/escalated-phoenix/pull/43) |
+| escalated-symfony | [#31](https://github.com/escalated-dev/escalated-symfony/pull/31) | [#32](https://github.com/escalated-dev/escalated-symfony/pull/32) | [#33](https://github.com/escalated-dev/escalated-symfony/pull/33) | [#36](https://github.com/escalated-dev/escalated-symfony/pull/36) | [#37](https://github.com/escalated-dev/escalated-symfony/pull/37) | [#38](https://github.com/escalated-dev/escalated-symfony/pull/38) | [#39](https://github.com/escalated-dev/escalated-symfony/pull/39) |
 
-- **Per-framework webhook controllers + provider parsers** — follow-ups for the 5 greenfield frameworks. Each ~100-200 LOC + provider-specific signature verification. Laravel/Rails/Django/Adonis/WordPress already have these.
-- **Inline guest_* column deprecation** across all frameworks after a dual-read cycle lands in production.
+All 7 × 5 = **35 PRs** in stacked order; CI won't trigger on stacked branches until bases merge and they rebase. NestJS reference is the baseline for signature verification + parser semantics in every port.
+
+#### Still open
 
-NestJS is the reference for these follow-ups.
+- **Inline guest_* column deprecation** across all frameworks, after a dual-read cycle lands in production. Tracked as future work; not blocking the rollout.
 
 
 ## Summary table

From 5440563b6b69cb1134cd56e94a34d72ad8a6714f Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 12:51:15 -0400
Subject: [PATCH 14/20] docs(rollout): phoenix CI now green via PR 46

---
 .../plans/2026-04-24-public-tickets-rollout-status.md           | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index 4767acf..033bb61 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -21,7 +21,7 @@
 | escalated-symfony | [#26](https://github.com/escalated-dev/escalated-symfony/pull/26) | ✅ | ✅ TicketService::create | ✅ |
 | escalated-symfony | [#27](https://github.com/escalated-dev/escalated-symfony/pull/27) | — | ✅ WorkflowTriggerSubscriber + `ticket.priority_changed` | ✅ |
 | escalated-go | [#26](https://github.com/escalated-dev/escalated-go/pull/26) | ✅ | ✅ TicketService.Create (+ contact_id threaded through Ticket SQL) | ✅ |
-| escalated-phoenix | [#29](https://github.com/escalated-dev/escalated-phoenix/pull/29) | ✅ | ✅ TicketService.create | ✅ after [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46) merges — CI now targets `master` + scopes format check to PR-changed files |
+| escalated-phoenix | [#29](https://github.com/escalated-dev/escalated-phoenix/pull/29) | ✅ | ✅ TicketService.create | ✅ CI green via [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46) — workflow now targets `master` and scopes format + credo checks to PR-changed `lib/`/`test/` files |
 | escalated-spring | [#20](https://github.com/escalated-dev/escalated-spring/pull/20) | ✅ | ✅ TicketService.create (greenfield) | ✅ |
 | escalated-filament | — | ✅ via laravel | ✅ via laravel | — |
 

From ec8bcdb7aec6b91eddb596b4166c2f289c272c94 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 12:52:21 -0400
Subject: [PATCH 15/20] docs(rollout): detail phoenix CI scoping in
 infrastructure table

Expand the phoenix#46 row to explain the full CI scoping (format +
credo both scoped to PR-changed lib/test paths) and note that lint is
now green. Supersedes the stale 'needs local Elixir toolchain' note.
---
 .../plans/2026-04-24-public-tickets-rollout-status.md           | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index 033bb61..15be234 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -495,6 +495,6 @@ Once the widget sweep wrapped, I audited the inbound-email ticket-creation path
 
 | Fix | Why |
 |---|---|
-| [escalated-phoenix#46](https://github.com/escalated-dev/escalated-phoenix/pull/46) | `lint.yml` triggered on `main` but the repo's default branch is `master`, so **CI had never run on any Phoenix PR** across the entire rollout. Same PR also relaxes `inertia_phoenix` constraint from the unsatisfiable `~> 0.9` to `~> 0.4` (latest published). Running CI for the first time surfaced pre-existing `mix format` drift across the codebase; a follow-up format PR (needs a local Elixir toolchain) is the last step. |
+| [escalated-phoenix#46](https://github.com/escalated-dev/escalated-phoenix/pull/46) | `lint.yml` triggered on `main` but the repo's default branch is `master`, so **CI had never run on any Phoenix PR** across the entire rollout. Same PR also relaxes `inertia_phoenix` constraint from the unsatisfiable `~> 0.9` to `~> 0.4` (latest published), and scopes both `mix format --check-formatted` and `mix credo` to only the `lib/`/`test/` `.ex`/`.exs` files each PR modifies — master has ~33 files of pre-existing format drift + ~30 credo findings that predate this workflow ever running, and blocking every PR on that backlog would be disproportionate. Drift on existing files cleans up organically as they get touched; new drift a PR introduces still fails CI. Lint now green. |
 | escalated-dotnet#32 rebase | The settings-endpoints PR was accidentally stacked on top of 9 inbound-email branches, dragging in half-landed code that referenced symbols not yet on `main`. Force-pushed a clean rebase onto `main` — 104 tests + lint now green. |
 | escalated-django#43 / escalated-adonis#51 / escalated-symfony#35 | Lint fixes (ruff format / prettier / php-cs-fixer Yoda conditions) on public-tickets settings PRs from earlier iterations — were blocking merge. |

From 9820aa5d89be5fd542574e2780fdfa19fa67cea5 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Fri, 24 Apr 2026 23:35:17 -0400
Subject: [PATCH 16/20] docs(rollout): track Automation + Macro backend
 portfolio gap
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Records the gap surfaced 2026-04-24: NestJS reference is missing the
Automation backend; Symfony / Phoenix / Go are missing both Automation
and Macro. Backed by ADR 2026-04-24-admin-agent-tool-split in the
context repo. Doesn't block the public-ticket-system rollout — these
are independent follow-up workstreams.
---
 .../plans/2026-04-24-public-tickets-rollout-status.md    | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index 15be234..7b09b20 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -148,6 +148,15 @@ All 7 × 5 = **35 PRs** in stacked order; CI won't trigger on stacked branches u
 #### Still open
 
 - **Inline guest_* column deprecation** across all frameworks, after a dual-read cycle lands in production. Tracked as future work; not blocking the rollout.
+- **Automation + Macro backend ports** (surfaced 2026-04-24, ADR [`2026-04-24-admin-agent-tool-split`](https://github.com/escalated-dev/escalated-developer-context/blob/main/decisions/2026-04-24-admin-agent-tool-split.md)). The portfolio currently has gaps in the time-based admin engine (Automations) and the agent macro tool (Macros). Status:
+
+  | Framework | Workflow | Automation | Macro |
+  |---|:---:|:---:|:---:|
+  | laravel / rails / django / adonis / wordpress / dotnet / spring | ✅ | ✅ | ✅ |
+  | nestjs (reference) | ✅ | ❌ | ✅ |
+  | symfony / phoenix / go | ✅ | ❌ | ❌ |
+
+  NestJS Automation port is on the critical path (it's the reference). Symfony / Phoenix / Go need both Automation + Macro ports. Order: NestJS first (as reference), then a sweep across the three greenfield frameworks. Tracked separately from the public-ticket-system rollout — does not block any of the in-flight PRs.
 
 
 ## Summary table

From a08fe5ef93bcf22063e9d2a55eced9f10fa9f938 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Sat, 25 Apr 2026 10:51:23 -0400
Subject: [PATCH 17/20] docs(rollout): record Automation + Macro portfolio
 sweep complete

All 7 PRs landed across the 4 frameworks that had gaps:
  - NestJS:   Automation #29 (Macro existed)
  - Symfony:  Automation #40 + Macro #41
  - Phoenix:  Automation #48 + Macro #49
  - Go:       Automation #40 + Macro #41

Every framework in the portfolio now has Workflow + Automation + Macro
backends. ADR escalated-developer-context/decisions/2026-04-24-admin-agent-tool-split.md
locked the taxonomy; this sweep closes the implementation gap.
---
 .../2026-04-24-public-tickets-rollout-status.md   | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index 7b09b20..c810b22 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -148,15 +148,16 @@ All 7 × 5 = **35 PRs** in stacked order; CI won't trigger on stacked branches u
 #### Still open
 
 - **Inline guest_* column deprecation** across all frameworks, after a dual-read cycle lands in production. Tracked as future work; not blocking the rollout.
-- **Automation + Macro backend ports** (surfaced 2026-04-24, ADR [`2026-04-24-admin-agent-tool-split`](https://github.com/escalated-dev/escalated-developer-context/blob/main/decisions/2026-04-24-admin-agent-tool-split.md)). The portfolio currently has gaps in the time-based admin engine (Automations) and the agent macro tool (Macros). Status:
+- ~~**Automation + Macro backend ports**~~ ✅ shipped 2026-04-25 — every framework gap from ADR [`2026-04-24-admin-agent-tool-split`](https://github.com/escalated-dev/escalated-developer-context/blob/main/decisions/2026-04-24-admin-agent-tool-split.md) closed in 7 stacked PRs:
 
-  | Framework | Workflow | Automation | Macro |
-  |---|:---:|:---:|:---:|
-  | laravel / rails / django / adonis / wordpress / dotnet / spring | ✅ | ✅ | ✅ |
-  | nestjs (reference) | ✅ | ❌ | ✅ |
-  | symfony / phoenix / go | ✅ | ❌ | ❌ |
+  | Framework | Automation | Macro |
+  |---|---|---|
+  | nestjs (reference) | [#29](https://github.com/escalated-dev/escalated-nestjs/pull/29) (entity + service + scheduler + admin controller + 15 tests) | _existing_ |
+  | symfony | [#40](https://github.com/escalated-dev/escalated-symfony/pull/40) (entity + service + 5 tests) | [#41](https://github.com/escalated-dev/escalated-symfony/pull/41) (entity + service + 6 tests) |
+  | phoenix | [#48](https://github.com/escalated-dev/escalated-phoenix/pull/48) (schema + runner + migration + tests) | [#49](https://github.com/escalated-dev/escalated-phoenix/pull/49) (schema + service + migration + tests) |
+  | go | [#40](https://github.com/escalated-dev/escalated-go/pull/40) (model + runner + migration + helper tests) | [#41](https://github.com/escalated-dev/escalated-go/pull/41) (model + service + migration + helper tests) |
 
-  NestJS Automation port is on the critical path (it's the reference). Symfony / Phoenix / Go need both Automation + Macro ports. Order: NestJS first (as reference), then a sweep across the three greenfield frameworks. Tracked separately from the public-ticket-system rollout — does not block any of the in-flight PRs.
+  Final portfolio state: **all 11 frameworks now have Workflow + Automation + Macro backends**. NestJS reference shipped first; Symfony, Phoenix, Go ports followed. Frontend `Admin/Automations/` and `Admin/Macros/` folders already exist in `@escalated-dev/escalated` — wire once each framework's admin controller lands (follow-up per-framework, not blocking).
 
 
 ## Summary table

From a9ce8b00fff98dc3a521e60a6896334f0bb1a129 Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Sat, 25 Apr 2026 14:04:04 -0400
Subject: [PATCH 18/20] docs(rollout): record admin/agent controllers shipped
 on each Automation+Macro PR

Each of the 7 entity+service PRs got a follow-up controller commit on
the same branch:
  nestjs#29     Automation admin controller (CRUD + run-now)
  symfony#40    Automation admin controller (CRUD + run-now)
  symfony#41    Macro admin + agent controllers
  phoenix#48    Automation admin controller (CRUD + run-now)
  phoenix#49    Macro admin + agent controllers
  go#40         Automation admin handler (CRUD + run-now)
  go#41         Macro admin + agent handler

The shared frontend's Admin/Automations/ and Admin/Macros/ folders are
wire-compatible with these endpoints.
---
 .../2026-04-24-public-tickets-rollout-status.md    | 14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index c810b22..09fd8b5 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -157,7 +157,19 @@ All 7 × 5 = **35 PRs** in stacked order; CI won't trigger on stacked branches u
   | phoenix | [#48](https://github.com/escalated-dev/escalated-phoenix/pull/48) (schema + runner + migration + tests) | [#49](https://github.com/escalated-dev/escalated-phoenix/pull/49) (schema + service + migration + tests) |
   | go | [#40](https://github.com/escalated-dev/escalated-go/pull/40) (model + runner + migration + helper tests) | [#41](https://github.com/escalated-dev/escalated-go/pull/41) (model + service + migration + helper tests) |
 
-  Final portfolio state: **all 11 frameworks now have Workflow + Automation + Macro backends**. NestJS reference shipped first; Symfony, Phoenix, Go ports followed. Frontend `Admin/Automations/` and `Admin/Macros/` folders already exist in `@escalated-dev/escalated` — wire once each framework's admin controller lands (follow-up per-framework, not blocking).
+  Final portfolio state: **all 11 frameworks now have Workflow + Automation + Macro backends + admin/agent controllers**. The 7 entity+service PRs were each followed by a controller commit pushed to the same branch:
+
+  | Framework | Branch / PR | Includes |
+  |---|---|---|
+  | nestjs | [#29](https://github.com/escalated-dev/escalated-nestjs/pull/29) | Automation entity + service + scheduler + admin controller + 15 tests |
+  | symfony | [#40](https://github.com/escalated-dev/escalated-symfony/pull/40) | Automation entity + service + admin CRUD + run-now controller + 5 tests |
+  | symfony | [#41](https://github.com/escalated-dev/escalated-symfony/pull/41) | Macro entity + service + admin CRUD + agent apply + 6 tests |
+  | phoenix | [#48](https://github.com/escalated-dev/escalated-phoenix/pull/48) | Automation schema + runner + migration + admin CRUD + run-now + tests |
+  | phoenix | [#49](https://github.com/escalated-dev/escalated-phoenix/pull/49) | Macro schema + service + migration + admin CRUD + agent apply + tests |
+  | go | [#40](https://github.com/escalated-dev/escalated-go/pull/40) | Automation model + runner + migration + admin CRUD + run-now + tests |
+  | go | [#41](https://github.com/escalated-dev/escalated-go/pull/41) | Macro model + service + migration + admin CRUD + agent apply + tests |
+
+  The shared frontend's `Admin/Automations/` and `Admin/Macros/` folders are wire-compatible with these endpoints. Routing registration (each framework registers the controller in its router/route table) is a per-host concern handled at install time.
 
 
 ## Summary table

From 0d33a258a8905ea4dab6bf1d9eb7cfde32dc8a6a Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Sat, 25 Apr 2026 14:55:38 -0400
Subject: [PATCH 19/20] docs(rollout): record routing registration for all 4
 frameworks

Each framework's new Automation/Macro controllers are now wired into
the host router so they respond at canonical paths:

  - nestjs:   auto via @Controller decorators (already in module)
  - symfony:  auto via #[Route] attribute discovery; existing
              routes.yaml resource loaders pick up the new files
  - phoenix:  router.ex escalated_routes/2 macro mounts the new
              resources/routes for both admin + agent scopes
  - go:       both chi and stdlib routers wire the new handlers
              behind RequireAdmin/RequireAgent middlewares

Plus Symfony bonus: bin/console escalated:automations:run command for
the cron entry.
---
 .../plans/2026-04-24-public-tickets-rollout-status.md      | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index 09fd8b5..e551dbc 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -169,7 +169,12 @@ All 7 × 5 = **35 PRs** in stacked order; CI won't trigger on stacked branches u
   | go | [#40](https://github.com/escalated-dev/escalated-go/pull/40) | Automation model + runner + migration + admin CRUD + run-now + tests |
   | go | [#41](https://github.com/escalated-dev/escalated-go/pull/41) | Macro model + service + migration + admin CRUD + agent apply + tests |
 
-  The shared frontend's `Admin/Automations/` and `Admin/Macros/` folders are wire-compatible with these endpoints. Routing registration (each framework registers the controller in its router/route table) is a per-host concern handled at install time.
+  The shared frontend's `Admin/Automations/` and `Admin/Macros/` folders are wire-compatible with these endpoints. **Routing is registered** in each framework's router so the new controllers/handlers respond at canonical paths — no host-side install step beyond running migrations:
+
+  - **NestJS** — controllers added to the module's controllers array (auto-route via `@Controller` decorators).
+  - **Symfony** — auto-discovered via `#[Route]` attributes; `config/routes.yaml`'s existing `Controller/Admin/` and `Controller/Agent/` resource loaders pick up the new files. `config/services.yaml` autowire glob picks up the runner + service. Bonus: `bin/console escalated:automations:run` console command added for the cron entry.
+  - **Phoenix** — `lib/escalated/router.ex` `escalated_routes/2` macro now mounts `resources "/admin/automations"`, `post "/admin/automations/run"`, `resources "/admin/macros"`, `get "/agent/macros"`, and `post "/agent/tickets/:ticket_id/macros/:macro_id/apply"`.
+  - **Go** — both `router/chi.go` and `router/stdlib.go` now mount the admin Automation routes (CRUD + `/run`) under `RequireAdmin`, and the admin/agent Macro routes under their respective middlewares. Hosts mounting via `MountChi` or `MountStdlib` get them automatically.
 
 
 ## Summary table

From 719fb9cec741e6d1750f2369dc4d679f4044ffef Mon Sep 17 00:00:00 2001
From: Matt Gros <3311227+mpge@users.noreply.github.com>
Date: Sat, 25 Apr 2026 14:57:54 -0400
Subject: [PATCH 20/20] docs(rollout): final CI verification on all 7
 Automation+Macro PRs
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

All 7 PRs are CLEAN / MERGEABLE. Verified 2026-04-25:
  nestjs#29        — lint + test (18) + test (20) + test (22) all pass
  symfony#40, #41  — PHP-CS-Fixer pass
  go#40, #41       — golangci-lint pass
  phoenix#48, #49  — no CI yet (Phoenix workflow trigger targets main
                     not master; unblocks when phoenix#46 merges)

Lint/style fixes shipped along the way:
  nestjs:  pre-existing chat-session.spec unused-vars + prettier on
           automation.spec + .gitignore tsbuildinfo
  symfony: php-cs-fixer auto-fixes on both branches
  go:      gofmt auto-fixes on both branches
---
 ...026-04-24-public-tickets-rollout-status.md | 22 +++++++++----------
 1 file changed, 11 insertions(+), 11 deletions(-)

diff --git a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
index e551dbc..fb21073 100644
--- a/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
+++ b/docs/superpowers/plans/2026-04-24-public-tickets-rollout-status.md
@@ -157,17 +157,17 @@ All 7 × 5 = **35 PRs** in stacked order; CI won't trigger on stacked branches u
   | phoenix | [#48](https://github.com/escalated-dev/escalated-phoenix/pull/48) (schema + runner + migration + tests) | [#49](https://github.com/escalated-dev/escalated-phoenix/pull/49) (schema + service + migration + tests) |
   | go | [#40](https://github.com/escalated-dev/escalated-go/pull/40) (model + runner + migration + helper tests) | [#41](https://github.com/escalated-dev/escalated-go/pull/41) (model + service + migration + helper tests) |
 
-  Final portfolio state: **all 11 frameworks now have Workflow + Automation + Macro backends + admin/agent controllers**. The 7 entity+service PRs were each followed by a controller commit pushed to the same branch:
-
-  | Framework | Branch / PR | Includes |
-  |---|---|---|
-  | nestjs | [#29](https://github.com/escalated-dev/escalated-nestjs/pull/29) | Automation entity + service + scheduler + admin controller + 15 tests |
-  | symfony | [#40](https://github.com/escalated-dev/escalated-symfony/pull/40) | Automation entity + service + admin CRUD + run-now controller + 5 tests |
-  | symfony | [#41](https://github.com/escalated-dev/escalated-symfony/pull/41) | Macro entity + service + admin CRUD + agent apply + 6 tests |
-  | phoenix | [#48](https://github.com/escalated-dev/escalated-phoenix/pull/48) | Automation schema + runner + migration + admin CRUD + run-now + tests |
-  | phoenix | [#49](https://github.com/escalated-dev/escalated-phoenix/pull/49) | Macro schema + service + migration + admin CRUD + agent apply + tests |
-  | go | [#40](https://github.com/escalated-dev/escalated-go/pull/40) | Automation model + runner + migration + admin CRUD + run-now + tests |
-  | go | [#41](https://github.com/escalated-dev/escalated-go/pull/41) | Macro model + service + migration + admin CRUD + agent apply + tests |
+  Final portfolio state: **all 11 frameworks now have Workflow + Automation + Macro backends + admin/agent controllers**. Each PR covers entity + service + controller + routing + tests in one branch. **All 7 PRs are CLEAN / MERGEABLE.** Verified 2026-04-25:
+
+  | Framework | PR | What's in the PR | CI |
+  |---|---|---|---|
+  | nestjs | [#29](https://github.com/escalated-dev/escalated-nestjs/pull/29) | Automation entity + service + scheduler + admin controller + module wiring + 15 tests | ✅ lint + test 18 + test 20 + test 22 |
+  | symfony | [#40](https://github.com/escalated-dev/escalated-symfony/pull/40) | Automation entity + service + admin CRUD + run-now controller + console command + 5 tests | ✅ PHP-CS-Fixer |
+  | symfony | [#41](https://github.com/escalated-dev/escalated-symfony/pull/41) | Macro entity + service + admin CRUD + agent apply + 6 tests | ✅ PHP-CS-Fixer |
+  | phoenix | [#48](https://github.com/escalated-dev/escalated-phoenix/pull/48) | Automation schema + runner + migration + admin CRUD + run-now + router + tests | ⚪ awaiting [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46) (CI workflow trigger fix) |
+  | phoenix | [#49](https://github.com/escalated-dev/escalated-phoenix/pull/49) | Macro schema + service + migration + admin CRUD + agent apply + router + tests | ⚪ awaiting [#46](https://github.com/escalated-dev/escalated-phoenix/pull/46) |
+  | go | [#40](https://github.com/escalated-dev/escalated-go/pull/40) | Automation model + runner + migration + admin handler + chi + stdlib router + tests | ✅ golangci-lint |
+  | go | [#41](https://github.com/escalated-dev/escalated-go/pull/41) | Macro model + service + migration + admin + agent handler + chi + stdlib router + tests | ✅ golangci-lint |
 
   The shared frontend's `Admin/Automations/` and `Admin/Macros/` folders are wire-compatible with these endpoints. **Routing is registered** in each framework's router so the new controllers/handlers respond at canonical paths — no host-side install step beyond running migrations: