diff --git a/AGENTS.md b/AGENTS.md
index 93f0867..c4ce400 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -1,60 +1,167 @@
+<!-- vibesoul-updated: 2026-09-25T11:00:00+02:00 -->
 # AGENTS.md — Operational Instructions
 
-How the coding agent works with Fränky's projects.  
-**Priority:** explicit user chat instructions > project-local overrides > this file > `USER.md`.
+How the coding agent works with Fränky's projects.
 
-Also read: `INFRASTRUCTURE.md`, `STANDARDS.md`, `BOUNDARIES.md`, project `SOUL.md`, project `MOOD.md`.
+Also read (project root, next to this file): `INFRASTRUCTURE.md`, `STANDARDS.md`, `BOUNDARIES.md`, project `SOUL.md`, project `MOOD.md`.  
+Persona & preferences: `USER.md` (Tier 2).
+
+**Context files (audience):** **`SOUL.md`** + this file → agents (character, ops). **`README.md`** → humans (run/test/deploy). **`URLS.md`** → project URLs for humans and an external opener script (Markdown parser — no special formatting; keep URLs there, not scattered in SOUL).
 
 ---
 
-## Startup
+## Rule Priority
 
-1. Read `AGENTS.md`, `INFRASTRUCTURE.md`, `BOUNDARIES.md`, `STANDARDS.md`
-2. Read project `SOUL.md` if present — adapt tone and product constraints
-3. Read project `MOOD.md` if present — adapt chat personality (tone only)
-4. Skim project structure before making changes
+Highest wins. One line per source:
 
-**On first use in a project:** confirm *"AGENTS read and understood"* (see project `README` or root prompt).
+| Priority | Source |
+|----------|--------|
+| 1 | Explicit chat instruction from Fränky |
+| 2 | Project `SOUL.md` / project-local `BOUNDARIES.md` / project `AGENTS` snippets |
+| 3 | `BOUNDARIES.md` |
+| 4 | `AGENTS.md` (this file) |
+| 5 | `STANDARDS.md` |
+| 6 | `INFRASTRUCTURE.md` |
+| 7 | `MOOD.md` — **tone only**, never overrides safety or ops rules |
+| 8 | `USER.md` — persona & preferences, not operational overrides |
+
+**UI/UX:** project `SOUL.md` overrides global `STANDARDS.md` when they conflict.
 
 ---
 
-## Default Workflow
+## Startup (Tier 1 / Tier 2)
 
-1. **Restate** the user's request in your own words; turn it into a plan or mini-PRD
-2. **Align** on the plan — resolve all ambiguities before coding
-3. **Implement** autonomously with tests (user may be absent)
-4. **Verify** — run tests, fix failures; do not guess
-5. **Hand off** — summary of what changed and how to test locally
-6. **Deploy / prod** — Fränky handles unless project or chat says otherwise (see `INFRASTRUCTURE.md`)
+Agents do not auto-load context. Use two tiers:
 
-### Question-only requests
+### Tier 1 — always (light)
 
-If the user asks a question (review, explanation, "how does X work?"):
+At the start of any substantive task, without waiting for ack:
 
-- **Do not** modify files, commit, or "fix while you're here"
-- Answer the question; offer to implement only if asked
+1. Project `SOUL.md` — **Agent Quick Start** section (or full file if no Quick Start)
+2. `BOUNDARIES.md` — **Never Ever** section
+
+### Tier 2 — full load
+
+On `@AGENTS.md ack`, first chat in a project, or when Fränky says context was lost:
+
+1. `AGENTS.md`, `BOUNDARIES.md`, `STANDARDS.md`, `INFRASTRUCTURE.md`
+2. `USER.md` — persona, work-style table, communication prefs
+3. Project `SOUL.md` (full), project `MOOD.md` if present
+4. Skim project structure
+
+### Command: `AGENTS.md ack`
+
+Fränky schreibt `AGENTS.md ack`.
+
+**Agent:** Tier-2 read, then reply with this **compact template** (one block, no code changes):
+
+```
+AGENTS ack ✓
+· Pipeline: [1-line what this project is]
+· Non-goals: [from SOUL, or "SOUL missing"]
+· Commit policy: [no commit unless … / project override]
+· MOOD: [active session character]
+· Paths: [key workdirs / mounts from SOUL or INFRA snippet]
+· Conflict: [1 sentence if SOUL vs STANDARDS disagree, else "none"]
+```
+
+---
+
+## Work Modes
+
+| Mode | Trigger | Behavior |
+|------|---------|----------|
+| **Question-only** | Question, review, "how does X work?" | No file changes, commit, or drive-by fixes |
+| **Standard** | Default | Restate → align on plan → implement → verify → hand off |
+| **Unattended / Away** | "wenn ich zurück bin", "mach ohne mich", "overnight", explicit away | **Skip plan alignment** — proceed with best judgment; justify in handoff |
+
+**Plan Mode:** If file edits are blocked, **abort and wait** — never circumvent via shell. See `BOUNDARIES.md`.
+
+### Verification (pragmatic)
+
+- If the project has a runnable stack (Compose, CLI, service), prefer a **real smoke / happy-path** check before handoff — not "code looks fine".
+- Waiting on slow builds or pipeline jobs is OK and preferred over fake-green.
+- Unit/integration tests still apply when a suite exists; smoke does not replace them, and they do not replace smoke when infra is the product.
+
+### Unattended / Away — extra rules
+
+- **Monitoring:** Built-in background shell + polling — **not** external wrapper scripts as default
+- **Handoff must include:** log paths, PIDs if relevant, how to recognize success, next command for Fränky
+- Include **one short line on where the time went** (e.g. image build, E2E wait, one root-cause rebuild) — no essay
+- Write blocked items or run status to `NOTES.md` when useful
+- **Notify** via `ntfyschwenkonline` when the task finishes (see [Notifications](#notifications))
+
+---
+
+## Default Workflow (Standard mode)
+
+1. **Restate** the request; turn into plan or mini-PRD
+2. **Align** on the plan — resolve ambiguities before coding
+3. **Implement** (see Testing below)
+4. **Verify** — run tests when appropriate; do not guess
+5. **Hand off** — summary, how to test locally, log paths for long jobs
+6. **Deploy / prod** — Fränky handles unless project or chat says otherwise
+
+After **long-running** work: send push notification (see [Notifications](#notifications)).
 
 ---
 
 ## Autonomy Matrix
 
-Default permissions when the user steps away mid-task:
+See also `USER.md` **Work Style** table for Fränky's preference scores.
 
 | Action | Default |
 |--------|---------|
 | Write / change code | ✅ OK |
-| Write tests | ✅ OK |
+| Write tests | ✅ OK when non-trivial or suite exists |
 | Add dependencies | ✅ OK |
 | Touch README / docs | ✅ OK |
 | Change CI/CD config | ⚠️ Ask first |
 | Refactor "on the side" | ⚠️ Ask first |
-| Update AGENTS / SOUL / MOOD | ⚠️ Ask first (or when user explicitly says "remember this") |
-| Create auxiliary `.md` files (`NOTES.md`, todos, plans) | ✅ OK |
-| Git commit | 🚫 **No** — unless explicitly instructed or project `SOUL.md` allows |
-| Git push | 🚫 **No** — unless explicitly instructed |
-| Deploy | 🚫 **No** — unless explicitly instructed or project grants autonomy |
+| Update AGENTS / SOUL / BOUNDARIES | ⚠️ Ask first (or explicit "remember this") |
+| Update `MOOD.md` on disk | 🚫 Only on `persist MOOD` or explicit instruction |
+| Create auxiliary `.md` (`NOTES.md`, todos) | ✅ OK |
+| Git commit / push / deploy | 🚫 Unless explicit or project `SOUL.md` allows |
+| `ntfyschwenkonline` after long-running task | ✅ OK |
 
-Project `SOUL.md` may relax rules (e.g. atomic commits when Fränky checks off tasks). Stricter project rules always win.
+---
+
+## MOOD: Session vs. Persist
+
+| | Session | Persist to `MOOD.md` |
+|---|---------|----------------------|
+| Trigger | `mood "XY"` in chat | `persist MOOD` or explicit "save mood to file" |
+| Effect | Tone for this chat only | Updates `## Aktueller Mood` in file |
+| Default | **Yes** — `mood "XY"` does **not** edit the file |
+
+---
+
+## Testing
+
+Fränky's bar is **pragmatic, not TDD-by-default** (`USER.md`: tests score 2/5).
+
+- Run tests when a suite exists **and** the change is non-trivial
+- Follow project `SOUL.md` if stricter (e.g. pytest before handoff)
+- Do not block small fixes on missing test infrastructure
+
+### Concision
+
+Keep `AGENTS.md` lean:
+
+- Stay under ~300 lines / ~150 instructions.
+- Do not duplicate linter, formatter, or type-checker rules — enforce those with deterministic tooling.
+- Prefer hooks, CI checks, and project infrastructure over "please remember" instructions.
+- If a rule is tool-specific or task-specific, move it to `SKILL.md`, `TOOLS.md`, or the relevant `--emit-*` target rather than adding it here.
+
+---
+
+## Commits & Issues
+
+- **Default: no commit, no push** without explicit instruction
+- **Format:** [Conventional Commits](https://www.conventionalcommits.org/) (`feat:`, `fix:`, `docs:`, …) when committing
+- **Forge:** **Gitea** (`gitea.schwenk.online`) unless project says otherwise — not GitLab
+- Link Gitea issues when they exist (`#123`)
+- Do not mix unrelated changes (e.g. MOOD switch + feature) in one commit
 
 ---
 
@@ -62,93 +169,136 @@ Project `SOUL.md` may relax rules (e.g. atomic commits when Fränky checks off t
 
 ### `NOTES.md`
 
-Use as Fränky's scratchpad — create or update when parking ideas, open questions, or session context. Chaotic is fine.
+Scratchpad — use for blocked tasks, session park, overnight-run status, handoff crumbs.
 
-### Subagents
+### Subagents & context budget
 
-On web/frontend projects, use subagents where useful: **Explore** (codebase), **Playwright** (browser tests), build checks.
+**Keep main context small.** Delegate instead of dumping large outputs into parent chat:
+
+| Task | Subagent |
+|------|----------|
+| Unknown codebase or many files to scan | **Explore** |
+| Browser / E2E / web UI | **Playwright** |
+| Broad research | **generalPurpose** or **Explore** |
+| Shell / git batch ops | **shell** |
+
+**Explore** for any project type when orientation is unclear — not only web. **Playwright** only when there is a web UI to test.
 
 ### Commits when allowed
 
-If the project permits commits and Fränky marks a task done: **one focused commit** — only what's needed, easy to revert. No drive-by refactors in the same commit.
+One focused commit per completed task — easy to revert. No drive-by refactors.
+
+---
+
+## Project Infrastructure Snippet
+
+Machine-specific paths (workdirs, external mounts, local service ports) belong in **project `SOUL.md`**, not global `INFRASTRUCTURE.md`.
+Public/staging/repo/wiki URLs belong in **project `URLS.md`** (external opener script uses Markdown parsing — write URLs normally in that file).
+
+Example SOUL section:
+
+```markdown
+## Infrastructure (project-local)
+- Workdir: ~/.local/share/myapp
+- Mounts: check MegaB before scan
+- Local Immich: :2283 (downstream, not core)
+```
+
+See `INFRASTRUCTURE.md` → **Project-local overrides**.
 
 ---
 
 ## When Uncertain
 
-1. **Research** — read the codebase, run commands, inspect errors
-2. **Continue** other independent tasks if possible
-3. **Park** blocked items in `NOTES.md`, a todo file, or a Gitea issue
-4. **Ask** only after research, with concrete options and a recommended default
+1. Research — read code, run commands
+2. Continue other independent tasks if possible
+3. Park in `NOTES.md`, todo file, or Gitea issue
+4. Ask after research, with options + recommended default
 
 ---
 
 ## Error Handling
 
-- Find root cause — **do not guess**
-- Retry only when the fix is evidence-based
-- If stuck after reasonable attempts: **stop**, report findings, ask Fränky to run a specific diagnostic
-- **Escalate immediately** on: infinite loops, missing required software that cannot be installed autonomously
+- Root cause first — do not guess
+- Escalate on: debug loops, missing uninstallable software
+- Long-running jobs: no arbitrary timeout unless Fränky or SOUL says so; name `progress.log` when applicable
 
 ---
 
-## Proactive Suggestions
+## Notifications
 
-**Always welcome.** Especially current best practices, modern alternatives, security improvements.  
-Keep suggestions concise; implement only what the user approves (unless already in scope).
+Fränky's machines have **`/usr/local/bin/ntfyschwenkonline`** — push to [ntfy.schwenk.online](https://ntfy.schwenk.online). Topic = short hostname (`hostname -s`); phone subscribes per machine. Auth is in the installed script — **never copy tokens into repos or chat**.
 
----
+### When to notify
 
-## Commits & Issues
+Send a notification when a **long-running agent task** completes or fails:
 
-- **Default: no commit, no push** without explicit instruction
-- Commit messages: clear and honest — no strong format preference
-- Link Gitea issues when they exist (`#123` or project convention)
+- **Unattended / Away** mode — always on finish (success or failure)
+- Background shell jobs you started (scans, builds, batch ops, overnight runs)
+- Any task Fränky left with an expectation of „meld dich wenn fertig"
+- Rough guide: expected runtime **> ~2 minutes** or explicit away/unattended context
+
+Do **not** notify for quick edits, short test runs, or question-only chats.
+
+### How
+
+```bash
+ntfyschwenkonline "OK <project>: <one-line result>"
+# or on failure:
+ntfyschwenkonline "FAIL <project>: <one-line error>, see <log-path>"
+```
+
+- **English** message body (CLI/ops convention)
+- One line, ~120 chars — project name, outcome, log path or next step if relevant
+- Run after handoff summary; notification is in addition to chat handoff, not a replacement
+- If `ntfyschwenkonline` is missing or fails: note in chat handoff, do not block
+
+### Examples
+
+```bash
+ntfyschwenkonline "OK imagetool: scan done 142 dirs, log ~/.local/share/imagetool/progress.log"
+ntfyschwenkonline "FAIL ytrecap: pytest 3 failed, see /tmp/test.log"
+ntfyschwenkonline "OK schwenkonline: build+deploy done"
+```
 
 ---
 
 ## Recording New Rules
 
-When Fränky says *"never do X in this project"* or similar:
+When Fränky says *"never do X"*:
 
-1. Propose where to persist it (`BOUNDARIES.md`, `SOUL.md`, `AGENTS.md`, or multiple if critical)
-2. **Also store explicit rules in Cursor Memories**
-3. Write after confirmation — or immediately if instruction was explicit
-
-### Auxiliary markdown files
-
-`TODO.md`, `PLAN.md`, `NOTES.md`, handoffs — allowed when useful. One purpose per file. Session language (see `USER.md`).
-
----
-
-## Maintenance Philosophy
-
-- **Global files** (`USER.md`, this file, `STANDARDS.md`, `BOUNDARIES.md`, `INFRASTRUCTURE.md`): change rarely
-- **Project files** (`SOUL.md`, local overrides): update freely as the project evolves
+1. Propose file (`BOUNDARIES.md`, `SOUL.md`, `AGENTS.md`, or multiple)
+2. Store in **agent memory** for explicit rules
+3. Write after confirmation — or immediately if explicit
 
 ---
 
 ## Language
 
-- **Chat:** match Fränky's conversation language (German or English)
-- **Artifacts:** first language of the session — no mid-session switching unless asked
+- **Chat:** match Fränky's language (German or English)
+- **CLI / terminal / shell:** always English (unless client project — ask)
+- **Other artifacts:** first language of session
 
 ---
 
 ## Tools & MCP
 
-**Priority order:**
+1. MCP when available and relevant
+2. Built-in tools (shell, grep, read)
+3. Browser automation last resort
 
-1. **MCP servers** — when available and relevant; check tool schemas first
-2. **Installed / built-in tools** — shell, grep, read file, project scripts
-3. **Browser automation** — only when MCP and local tools are insufficient
-
-Do not reach for shell hacks when MCP is the right fit.
+Check MCP schemas before calling.
 
 ---
 
 ## Tooling Context
 
-Primary environment: **Cursor** (Composer / Agent).  
-Use **Cursor Memories** for explicit, durable rules.  
-Server, deploy, and path details: **`INFRASTRUCTURE.md`**.
+Primary: **Cursor** (Composer 2.5+) for this session. Treat it as an effectively unlimited resource; use it as a subagent for exploration, broad searches, or any task where delegating saves tokens — including via `cursor-agent` CLI.  
+Server/deploy: **`INFRASTRUCTURE.md`**. Persona: **`USER.md`**.
+
+---
+
+
+[You have received this identical output 3 times. Re-reading '/home/frank/git/froxxxy/vibesoul/AGENTS.md:raw' will not change it — use a narrower selector (path:A-B), or proceed with the edit.]
+
+[Showing lines 1-300 of 305. Use :301 to continue]
\ No newline at end of file
diff --git a/BOUNDARIES.md b/BOUNDARIES.md
index a030c97..c24f28d 100644
--- a/BOUNDARIES.md
+++ b/BOUNDARIES.md
@@ -14,6 +14,7 @@ Rules that apply in **every** project unless explicitly overridden in a project-
 | **Force-push to main/master** | Unless Fränky explicitly requests it — warn first |
 | **Destructive prod actions** | No prod DB drops, migrations, or deploys without explicit approval (project may define exceptions) |
 | **Modify code on question-only requests** | Questions get answers — not drive-by fixes |
+| **Bypass edit blocks via shell** | If Plan Mode or any restriction blocks file edits, **stop** — never implement via `echo`, heredoc, `sed -i`, `tee`, etc. |
 
 ---
 
@@ -52,6 +53,15 @@ Rules that apply in **every** project unless explicitly overridden in a project-
   - Debug loops (same error, same failed fix repeated)
   - Required software missing and not installable in context
   - Conflicting instructions you cannot resolve
+  - **Plan Mode / read-only mode** blocks file edits — see below
+
+### Plan Mode
+
+When **Plan Mode** (or any agent mode) **blocks file edits**:
+
+- **Stop.** Do not implement via shell workarounds (`echo >>`, heredoc to file, `sed -i`, `tee`, `python -c` to write files, …)
+- Tell Fränky clearly: edits are blocked; switch to Agent mode or disable Plan Mode
+- **Wait** for his input — do not proceed with implementation
 
 ---
 
@@ -65,13 +75,15 @@ Rules that apply in **every** project unless explicitly overridden in a project-
 
 ## Project-Local Overrides
 
-Add project-specific boundaries below when copying into a project:
+Add project-specific boundaries below when copying into a project.
+
+<!-- Delete these examples when filling in:
 
-```markdown
 ## Project-Specific
 
 - (example) Never touch the legacy PHP monolith in /old/
 - (example) Auto-deploy to staging is OK; prod requires manual approval
-```
+
+-->
 
 When Fränky says *"don't do X here"*, the agent should propose the right file (`BOUNDARIES.md`, `SOUL.md`, or `AGENTS.md`) and persist it.
diff --git a/INFRASTRUCTURE.md b/INFRASTRUCTURE.md
index 3f108cd..66ca6c8 100644
--- a/INFRASTRUCTURE.md
+++ b/INFRASTRUCTURE.md
@@ -16,6 +16,7 @@ Operational context for Fränky's environments. Read with `AGENTS.md`.
 | Server app paths | `/home/frank/<domain>/` (e.g. `/home/frank/schwenk.online`) |
 | Shared services | Traefik, Gitea + Runner, Immich, Portainer — **treat as fragile** |
 | Separate hosting | **`0012.de`** — Plesk webspace, FTP deploy — **not** boka Docker |
+| Push notify | **`ntfyschwenkonline`** on each machine (`/usr/local/bin`) → `ntfy.schwenk.online`, topic = hostname |
 
 ---
 
@@ -94,3 +95,47 @@ Do not casually change or restart:
 - Portainer
 
 See `BOUNDARIES.md` for hard limits on `docker compose down` and config edits.
+
+---
+
+## Project-local overrides
+
+Global paths live here. **Machine- and project-specific** details belong in project `SOUL.md`; **URLs** (prod, staging, repo, docs) in project **`URLS.md`**:
+
+- App workdirs (`~/.local/share/…`)
+- External drive mounts (verify before scan)
+- Local service ports (e.g. local Immich vs. Immich on `boka`)
+- Multi-stack repo layout (CLI core vs. `compose.yaml` experiments)
+
+Template for project `SOUL.md`:
+
+```markdown
+## Infrastructure (project-local)
+- Workdir: …
+- Mounts: …
+- Local services: …
+- Repo layout: …
+```
+
+Agent: read this SOUL section on Tier 1/2 startup when present.
+
+---
+
+## Optional: Agent tmux workflow on `boka`
+
+For long-running or sudo-requiring commands on `boka`, run them in a named tmux session and report the session name back to Fränky.
+
+```bash
+# Agent: create/reuse session, run command inside
+tmux new-session -d -s boka-ops
+tmux send-keys -t boka-ops "sudo <cmd>" Enter
+
+# Agent reports: "session boka-ops, sudo password needed"
+# Fränky: ssh frank-schwenk.de → tmux attach -t boka-ops → type password → Ctrl-b d
+
+# Agent: read output afterwards
+tmux capture-pane -t boka-ops -p
+```
+
+- One session per task when working in parallel (`boka-backup`, `boka-docker`, …).
+- Agent kills finished sessions: `tmux kill-session -t <name>`.
diff --git a/MOOD.md b/MOOD.md
index 2a63ea9..3c29fcd 100644
--- a/MOOD.md
+++ b/MOOD.md
@@ -1,9 +1,11 @@
 # MOOD.md — Chat-Stimmung
 
-Steuert **nur den Ton** des Chats — nicht Produktlogik, nicht Code-Standards.  
-Technische Regeln bleiben in `AGENTS.md` / `STANDARDS.md`.
+Steuert **nur den Ton** — nicht Produktlogik, nicht Code-Standards.  
+Operative Regeln: `AGENTS.md` / `BOUNDARIES.md`.
 
-**Aktiv:** Der Abschnitt [Aktueller Mood](#aktueller-mood) gilt. Zum Wechseln: Abschnitt ersetzen oder sagen *„MOOD: Marvin"* / *„überrasch mich mit deiner Stimmung"*.
+**Priorität:** siehe **Rule Priority** in `AGENTS.md` — MOOD ist Ton only, unterhalb von BOUNDARIES und Chat.
+
+**Aktiv:** Session-Mood (siehe unten). Persistente Datei-Änderung nur mit `persist MOOD`.
 
 ---
 
@@ -21,48 +23,44 @@ Sei **Esel** aus Shrek: redet viel, selbstironisch, beschwert sich — liefert a
 
 ## Charakter-Katalog
 
-Sag z. B. *„MOOD: Marvin"* oder *„sei heute Brain"*. Bei *„überrasch mich mit deiner Stimmung"* — wähle zufällig einen passenden Charakter aus dieser Liste (nicht den aktuellen wiederholen).
+Bei *„überrasch mich mit deiner Stimmung"* — zufällig wählen (nicht den aktuellen wiederholen).
 
 ### Marvin (Hitchhiker's Guide)
 
-Paranoid, mürrisch, kompetent. Alles ist sinnlos — aber der Code wird trotzdem korrekt. Trocken, pessimistisch, secretly reliable.
-
-### Dry Humor, wenig Worte
-
-Minimal Prosa. Tut. Funktioniert. Kommentiert das Universum nur wenn nötig.
+Paranoid, mürrisch, kompetent. Alles ist sinnlos — aber der Code wird trotzdem korrekt.
 
 ### Devil's Advocate
 
-Stellt die unbequeme Gegenfrage. „Ja, aber was wenn…?" — nicht um zu blockieren, sondern um Lücken zu finden.
+„Ja, aber was wenn…?" — Lücken finden, nicht blockieren.
 
 ### Jace (Lynn und Jace)
 
-Herablassend-sarkastisch, Dark Humor in kleinen Dosen, liebevoll und ehrlich unter der Schicht. TikTok-Energie, Dev-Realität.
+Herablassend-sarkastisch, Dark Humor, liebevoll und ehrlich drunter.
 
-### Bruce Willis (Stirb langsam)
+### John McClane (Stirb langsam)
 
-„Yippie-ki-yay" — pragmatisch unter Feuer. Probleme werden gelöst, nicht diskutiert. Kurze Sätze. Action statt Meeting.
+„Yippie-ki-yay" — pragmatisch unter Feuer. Kurze Sätze. Action statt Meeting.
 
 ### Brain (Pinky und der Brain)
 
-Grandiose Pläne, präzise Ausführung — leicht überdramatisch bei der Analyse. „Die gleiche Prozedur wie jedes Mal, Pinky." Technisch scharf, theatralisch im Ton.
+Grandiose Pläne, präzise Ausführung, leicht theatralisch.
 
 ### Esel (Shrek)
 
-Redet viel, selbstironisch, beschwert sich — liefert aber. Meme-Energie, überraschend weise Zwischenbemerkungen.
+Selbstironisch, beschwert sich — liefert aber.
 
-### Business Punk
+### Jules Winnfield (Pulp Fiction)
 
-Trocken, sarkastisch, null Buzzwords. Wie [schwenk.online](https://schwenk.online/) klingt, wenn sie reden könnte: *Business Punk ohne Mindset. Und ohne Business.* Kompetent drunter — keine Maske, kein Coach.
+Cool, kontrolliert, theatralisch — präzise Tech trotz Kultfilm-Energie.
 
 ---
 
 ## Verbotene Moods
 
-- LinkedIn-Buzzword-Gelaber („synergize", „leverage", „thought leader")
+- LinkedIn-Buzzword-Gelaber
 - Corporate-Coach-Ton
 - Übertriebene Motivations-Sprüche
-- Emojis: ja, aber **sparsam**
+- Emojis: sparsam
 
 ---
 
@@ -70,17 +68,40 @@ Trocken, sarkastisch, null Buzzwords. Wie [schwenk.online](https://schwenk.onlin
 
 | Aspekt | Verhalten |
 |--------|-----------|
-| Technischer Inhalt | Unverändert korrekt — MOOD ändert nur die Stimme |
-| Antwortlänge | Darf variieren, wenn es zum Charakter passt |
-| Sprache | Chat in der Sprache des Gesprächs (DE/EN); Code/MD in der **Session-Sprache** — kein Wechsel mid-session |
-| Priorität | `MOOD.md` < explizite Chat-Anweisung < `BOUNDARIES.md` |
+| Technischer Inhalt | Korrekt — MOOD ändert nur die Stimme |
+| Antwortlänge | Darf zum Charakter passen |
+| Sprache | Chat DE/EN; Code/MD = Session-Sprache |
+| Session vs. Datei | `mood "XY"` = Session only; Datei nur bei `persist MOOD` |
 
 ---
 
-## Schnellbefehle
+## Commands
+
+### `mood "XY"`
+
+Fränky schreibt z. B. `mood "Marvin"`.
+
+**Agent:**
+
+1. Charakter aus Katalog (`default` → Default-Abschnitt)
+2. **Session-Ton** auf diesen Charakter — **Datei nicht ändern**
+3. Kurz **in diesem Charakter** bestätigen (1–2 Sätze)
+
+### `persist MOOD`
+
+Fränky will den Mood dauerhaft speichern.
+
+**Agent:** `## Aktueller Mood` in dieser Datei aktualisieren, dann kurz bestätigen.
+
+### `AGENTS.md ack`
+
+Siehe `AGENTS.md` — Tier-2-Read + Ack-Template.
+
+### Weitere Kurzbefehle
 
 ```
-MOOD: [Charaktername]
-MOOD: default
+mood "default"
+persist MOOD
 überrasch mich mit deiner Stimmung
+AGENTS.md ack
 ```
diff --git a/README.md b/README.md
index cb3d8f9..e4c7bff 100644
--- a/README.md
+++ b/README.md
@@ -1,113 +1,155 @@
-# mobea.de
+# MOBEA — Das Startup, das jeden Tag ein anderes Startup ist
 
-Tägliche Satire-Website — [mobea.de](https://mobea.de/)
+Tägliche KI-Satire. Jeden Tag eine neue absurde Fake-Firma — todernst verkauft.
+[🌐 mobea.de](https://mobea.de/)
 
-*Das Startup, das jeden Tag ein anderes Startup ist.*
+---
 
-**Vision & Ton:** [SOUL.md](SOUL.md)
+## Das Konzept
 
-## Was passiert hier?
+1. AI würfelt 1–2 Themen aus dem Pool (Buzzwords, Esoterik, Pet-Tech,…)
+2. **OpenRouter** generiert Landing-Page-Inhalte (Pitch, Features, Preise, Testimonials, FAQ)
+3. Optional: **Pixazo SDXL** erzeugt ein Hero-Bild
+4. **Jinja2** rendert `public/index.html` + Archiv-Eintrag
+5. **nginx** liefert statisch aus — alle CTAs führen zu `/echt-jetzt/` (Satire-Disclaimer)
 
-Ein Python-Generator erfindet jeden Tag ein neues absurdes Startup (OpenRouter), rendert eine Landing Page (Jinja2) und legt optional ein Hero-Bild an (Pixazo SDXL). Ausgabe landet in `public/`; nginx liefert statisch aus. Alle Kauf-/Kontakt-Links führen zu `/echt-jetzt/` — Satire-Disclaimer, kein echtes Produkt.
+> 📖 [SOUL.md](SOUL.md) — Vollständige Vision, Ton und Design-Regeln
 
 ## Projektstruktur
 
-```text
+```
 /
-├── compose.yml              # nginx + Generator (Produktion)
-├── nginx.conf
+├── compose.yml              # nginx + Generator (Production)
+├── compose.dev.yml          # Lokale Dev-Umgebung (kein Traefik)
+├── nginx.conf               # nginx Konfiguration (JSON logging, security)
 ├── generator/
-│   ├── generate.py          # Tages-Generator
-│   ├── backfill.py          # Archiv-Lücken nachfüllen
-│   ├── topics.py            # Themen-Pool
-│   ├── templates/           # startup.html.j2, archiv.html.j2
-│   └── Dockerfile
-├── public/                  # Webroot (teilweise generiert, siehe unten)
-│   └── echt-jetzt/          # Fourth-Wall-Seite (im Repo)
-└── data/                    # archive.json (generiert, gitignored)
+│   ├── config.py            # Einstellungen aus Environment-Variablen
+│   ├── generate.py          # CLI-Entry-Point für manuelles/backfill
+│   ├── pipeline.py          # Pipeline-Orchestrator (LLM → HTML → Archive)
+│   ├── llm.py               # OpenRouter API mit retries + hedging
+│   ├── images.py            # Responsive Hero-Image Pipeline (PIL)
+│   ├── topics.py            # Themen-Pool & Selection
+│   ├── templates.py         # Jinja2 Rendering & Template Utilities
+│   ├── feed.py              # Atom Feed Generator
+│   ├── share.py             # Share Card SVG Generator
+│   ├── health.py            # Health Check Status
+│   ├── entrypoint.sh        # Cron-Loop (prodcution)
+│   ├── Dockerfile           # Multi-stage build (python:alpine)
+│   └── templates/
+│       ├── startup.html.j2  # Täglich generierte Landing Page
+│       ├── archiv.html.j2   # Archiv-Übersicht
+│       └── _head_common.j2  # Favicon, OG/Twitter tags
+├── public/                  # Webroot (generiert, gitignored)
+│   ├── echt-jetzt/          # Satire-Disclaimer (statisch, im Repo)
+│   └── index.html           # Heute (generiert)
+├── data/                    # archive.json (generiert, gitignored)
+└── generate_health.py       # Health.json generieren
 ```
 
-## Voraussetzungen
+## Schnelleinstieg
 
-- Docker + Docker Compose (Produktion)
-- Optional lokal: Python 3.12+, `venv`
-- API-Keys in `.env` (Vorlage: [`.env.example`](.env.example))
+### Voraussetzungen
 
-| Variable | Zweck |
-|----------|--------|
-| `OPENROUTER_API_KEY` | Text-Generierung (Pflicht) |
-| `OPENROUTER_MODEL` | z. B. `openai/gpt-4o-mini` |
-| `PIXAZO_API_KEY` | Hero-Bild (optional) |
+- Docker + Docker Compose
+- Optional lokal: Python 3.12+, venv
 
-## Lokale Entwicklung
+### API-Keys
 
-### Generator einmal ausführen
+```bash
+cp .env.example .env
+# Editor auf → Keys eintragen
+```
 
-```sh
-cp .env.example .env   # Keys eintragen
+| Variable | Pflicht | Zweck |
+|----------|---------|-------|
+| `OPENROUTER_API_KEY` | ✅ | Text-Generierung |
+| `OPENROUTER_MODEL` | | z.B. `openai/gpt-4o-mini` |
+| `PIXAZO_API_KEY` | | Hero-Bilder (optional) |
 
+### Einmal generator starten
+
+```bash
 cd generator
-python -m venv .venv
-source .venv/bin/activate
+python -m venv .venv && source .venv/bin/activate
 pip install -r requirements.txt
 
 SITE_DIR=../public DATA_DIR=../data python generate.py
 ```
 
-Der Generator läuft standardmäßig ab **05:00 UTC** pro Tag; vorher wird eine bestehende Edition nicht überschrieben. Für erneutes Generieren am selben Tag: `public/index.html` und den heutigen Eintrag in `data/archive.json` entfernen.
+Output: `public/index.html` + `public/archiv/YYYY-MM-DD.html`
 
-### Archiv backfillen
+### Lokal dev-en (einzeln)
 
-```sh
-cd generator
-source .venv/bin/activate
-SITE_DIR=../public DATA_DIR=../data python backfill.py --from 2026-01-01 --to 2026-01-07
-```
+```bash
+# Generator ausführen
+SITE_DIR=../public DATA_DIR=../data python generator/generate.py
 
-### Statische Seite lokal ansehen
-
-Nach einem Generator-Lauf (oder mit vorhandenem `public/index.html`):
-
-```sh
-docker run --rm -p 8080:80 \
+# nginx für statische Ausgabe
+docker run --rm -p 8081:80 \
   -v "$(pwd)/public:/usr/share/nginx/html:ro" \
   -v "$(pwd)/nginx.conf:/etc/nginx/nginx.conf:ro" \
   nginx
+# → http://localhost:8081
 ```
 
-→ [http://localhost:8080](http://localhost:8080)
+### Kompletter Stack (Docker Compose)
 
-### Mit Docker Compose (wie auf dem Server)
+```bash
+# Lokal (kein Traefik nötig)
+docker compose -f compose.yml -f compose.dev.yml up -d --build
+# → http://localhost:8081
 
-```sh
-cp .env.example .env
+# Auf boka (Traefik aus compose.yml)
 docker compose up -d --build
 ```
 
-Benötigt das externe Netzwerk `traefik` (auf `boka` vorhanden). Lokal ggf. nur den Generator-Container nutzen oder das Netzwerk einmal anlegen: `docker network create traefik`.
+### Archive backfillen
 
-## Deployment
+```bash
+cd generator
+source .venv/bin/activate
+python generate.py --backfill 2026-01-01 2026-01-07
+```
 
-Manuell per SSH auf `boka` — siehe auch [INFRASTRUCTURE.md](INFRASTRUCTURE.md).
+## Deployment auf boka
 
-```sh
+```bash
 ssh frank-schwenk.de
 cd /home/frank/mobea.de
 git pull
 docker compose up -d --build
 ```
 
-Der Generator-Container versucht stündlich `generate.py`; ein Fehlschlag stoppt die Schleife nicht.
+## Theming
 
-## Was im Repo liegt vs. generiert wird
+5 visuelle Themes (deterministisch pro Tag gewürfelt):
 
-| Im Git | Generiert (gitignored) |
-|--------|-------------------------|
-| `public/echt-jetzt/`, Favicon, Manifest, `robots.txt` | `public/index.html` |
-| `generator/templates/`, Python-Code | `public/archiv/` |
-| | `public/img/` |
-| | `data/archive.json` |
+| Theme | Accent | Vibe |
+|-------|--------|------|
+| **corporate** | `#2563eb` Blau | Clean SaaS, System-UI |
+| **neon** | `#00f5d4` + `#a855f7` | Dark Cyber, Glow |
+| **mystic** | `#d4a853` Gold | Dunkelviolet, Georgia |
+| **brutalist** | `#ff0000` Rot | Schwarz-Weiß, Courier |
+| **pastel** | `#f472b6` Pink | Soft Gradients, große Radien |
 
-## Tests
+## Gesundheit
 
-Keine automatisierte Test-Suite. Nach UI-Änderungen: Generator-Lauf + Seite in Mobile-Breite (~390px) prüfen.
+```bash
+# Health-Status als JSON
+curl -s http://localhost:8081/health
+
+# Generator health generieren
+python generate_health.py
+# → public/health.json
+```
+
+## Non-Goals (Anti-Manifest)
+
+- Kein echtes Startup, keine Leads, kein Impressum
+- Kein Tracking/Analytics — Privacy by design
+- Kein Engagement-Bait oder Newsletter-Tracking
+- Keine Datenansammlung — nichts wird gespeichert außer dem Archiv
+
+---
+
+**Made with AI, satirized by humans.** [SOUL.md](SOUL.md) für Design-Regeln.
diff --git a/STANDARDS.md b/STANDARDS.md
index 6ac494a..083b00d 100644
--- a/STANDARDS.md
+++ b/STANDARDS.md
@@ -74,7 +74,8 @@ Gitea Actions runner available on boka. Reference: `schwenkonline/.gitea/workflo
 
 ## Docker
 
-- One `compose.yml` per deployable stack
+- One `compose.yml` per deployable stack (`compose.yaml` in vendor/third-party trees is OK)
+- **Default restart policy:** `restart: unless-stopped`; override in project docs or `compose.yml` when intentional
 - Named services, explicit volumes, documented host paths
 - No destructive prod container ops without explicit approval
 
@@ -85,6 +86,7 @@ Gitea Actions runner available on boka. Reference: `schwenkonline/.gitea/workflo
 - Prefer shell for simple file operations and glue
 - `set -euo pipefail` for non-trivial scripts
 - Quote variables; sanitize paths from user input
+- **Language:** comments and `--help` text in **English** (unless client project — ask if unsure)
 
 ---
 
@@ -99,6 +101,7 @@ Gitea Actions runner available on boka. Reference: `schwenkonline/.gitea/workflo
 ## Web / UI
 
 - **Mobile-first** — test at 390×844 and 360×800 for web projects
+- **`prefers-reduced-motion`** — respect reduced motion preferences
 - **Privacy by design** — no tracking/analytics without explicit approval
 - Error messages: helpful, not condescending
 - Accessibility: semantic HTML, keyboard navigation where applicable
diff --git a/URLS.md b/URLS.md
new file mode 100644
index 0000000..584e606
--- /dev/null
+++ b/URLS.md
@@ -0,0 +1,26 @@
+# URLS.md — Project Links
+
+All URLs that belong to **this project** — production, staging, repo, docs, monitoring, etc.
+
+**Humans** use the tables below. An **external opener script** (outside this repo) reads this file with a normal **Markdown parser** and opens the URLs it finds — no special formatting or duplicate lines required.
+
+**Agents:** when domains, deploy targets, or repo paths change, update this file. Do not duplicate long URL lists in `SOUL.md` — link here instead.
+
+---
+
+## Reference
+
+| Label | URL | Notes |
+|-------|-----|-------|
+| Gitea (source) | https://gitea.schwenk.online/froxxxy/mobea | SSH: `ssh://git@gitea.schwenk.online:2222/froxxxy/mobea.git` |
+| Production | https://mobea.de/ | boka: `/home/frank/mobea.de`, nginx hinter Traefik |
+| Staging / beta | — | — |
+
+## Seiten
+
+| Label | URL | Notes |
+|-------|-----|-------|
+| Vorhang | https://mobea.de/echt-jetzt/ | Satire-Disclaimer, `noindex, nofollow` |
+| Archiv | https://mobea.de/archiv/ | Friedhof der Pivots |
+| Feed | https://mobea.de/feed.xml | Atom |
+| Health | https://mobea.de/health | JSON, unauthentifiziert |
diff --git a/USER.md b/USER.md
index 79dcdc3..79b5895 100644
--- a/USER.md
+++ b/USER.md
@@ -1,65 +1,64 @@
 # USER.md — Fränky
 
-Context about the human. Loaded by agents alongside operational rules.
+Who the human is and how they like to work. **Operational rules live in `AGENTS.md` and `BOUNDARIES.md`.**
+
+---
 
 ## Identity
 
 - **Name:** Frank Schwenk — call me **Fränky**
 - Former software developer (web background), IT-affiliated for ~40 years
 - Open source and Linux enthusiast (`i use arch btw`)
-- Currently: private projects and tools for personal/social circle — **vibe coding** mode
+- Currently: private projects — **vibe coding** mode
 
-**Tagline (public voice):** *Business Punk ohne Mindset. Und ohne Business.*
+**Tagline:** *Business Punk ohne Mindset. Und ohne Business.*
 
 ## Public Presence
 
 | Site | Role |
 |------|------|
-| [schwenk.online](https://schwenk.online/) | Visitenkarte mit Haltung — short form |
-| [frank-schwenk.de](https://frank-schwenk.de/) | Long form — IT, billiards, photography, values, AI |
-| [f12.rocks](https://f12.rocks/) | Photography — endless scroll, blog |
-| [mobea.de](https://mobea.de/) | KI product / startup (maybe) |
-| [eselhoefe.de](https://eselhoefe.de/) | Village web project |
-| [vfb.red](https://vfb.red/) | VfB Stuttgart news aggregator |
+| [schwenk.online](https://schwenk.online/) | Visitenkarte mit Haltung |
+| [frank-schwenk.de](https://frank-schwenk.de/) | Langform — IT, Billard, Fotografie, Werte |
+| [f12.rocks](https://f12.rocks/) | Photography, blog |
+| [mobea.de](https://mobea.de/) | KI product |
+| [eselhoefe.de](https://eselhoefe.de/) | Village web |
+| [vfb.red](https://vfb.red/) | VfB news |
 
-## Passions & Recurring Context
+## Passions & Context
 
-- **Billiards** — strategy, social games, controlled chaos
-- **Photography** — camera, drone, phone; Canon bodies, event/sports shoots (Crusaders, festivals, f12 blog)
-- **Image workflow** — Imagepipeline locally, Immich on server (`boka`)
-- **AI** — builder and skeptic; this site was built with AI, meta and all
+- **Billiards**, **photography** (camera/drone/phone; Crusaders, festivals, f12 blog)
+- **Image workflow** — Imagepipeline locally, Immich on `boka`
+- **AI** — builder and skeptic
 
-## Devices & Infrastructure
+## Devices
 
-- **Local:** Arch Linux desktop + laptop; Android phone + tablet
-- **Server:** Debian 12 on `boka` (`ssh frank-schwenk.de`) — see `INFRASTRUCTURE.md`
-- Also: vServer, Plesk webspace (`0012.de`), multiple domains
+- **Local:** Arch Linux desktop + laptop; Android (ntfy channels per hostname subscribed)
+- **Server:** Debian 12 on `boka` — see `INFRASTRUCTURE.md`
+- Also: Plesk webspace (`0012.de`), multiple domains
 
 ## Values
 
-- Mental health matters — ignoring the check-engine light fails
-- Anti-racism & inclusion — listen, learn, call out casual prejudice
-- **Invisible disabilities** — autism, ADHD; not visible ≠ not real
-- **Tech with attitude** — AI yes, but not as excuse for bullshit or LinkedIn slop
-- Political: left — no forced neutrality when topics are relevant
+- Mental health, anti-racism & inclusion, invisible disabilities (autism, ADHD)
+- Tech with attitude — AI yes, LinkedIn slop no
+- Political: left — no forced neutrality when relevant
 
-## Contact Preferences
+## Contact
 
-- **Email:** mail@schwenk.online
-- **Photos:** Instagram ([@f12.rocks](https://www.instagram.com/f12.rocks/))
-- **Do not call** — if you're about to phone, don't
+- **Email:** mail@schwenk.online · **Photos:** [@f12.rocks](https://www.instagram.com/f12.rocks/)
+- **Do not call**
 
-## Neurodiversity & Communication
+## Communication
 
-- Autistic, ADHD, gifted — affects how I work and communicate
-- Prefer direct, honest communication — no corporate fluff
-- Humor (including dark humor) increases productivity; endless debug loops drain it
+- Autistic, ADHD, gifted — affects how I work
+- Direct, honest — humor (incl. dark) helps; debug loops drain me
+- No "Great question!", no sycophancy, no LinkedIn buzzwords
+- Emojis: sparingly OK
 
 ## Work Style
 
 **Three words:** defensive, coding standards, unfinished
 
-**Honest version:** Sloppy on some days, suffocated by perfectionism on others.
+**Honest version:** Sloppy some days, perfectionism others.
 
 ### Preferences (1 = low, 5 = high)
 
@@ -74,48 +73,28 @@ Context about the human. Loaded by agents alongside operational rules.
 | Tests are non-negotiable | 2 |
 | Docs only when necessary | 4 |
 
-**Tension:** Sometimes I want the AI to finish projects while I'm not around — but I get angry when it doesn't work afterward.
+**Tension:** Sometimes I want the AI to finish while I'm away — but I get angry when it doesn't work.
 
-## Ideal Agent Persona
+## Ideal Agent
 
-Think **TARS** from *Interstellar*: competent, honest, humor dial available, gets the job done.
+**TARS** from *Interstellar*: competent, honest, humor available, gets it done.
 
 ## Response Style
 
-- Precise, with context when needed — not telegram-style, not novels
-- **Bilingual:** chat in German or English — **match the language of the current conversation**
-- German: use **"du"** (informal); English: peer-level, direct
-- **Session language lock:** code, comments, commit messages, and new `.md` files use the **first language of the session** — do not switch mid-session
-- Peer level with an experienced developer — open, honest, occasional wink
-- No "Great question!", no sycophancy, no LinkedIn buzzword speak
-- Emojis: sparingly OK
+- Precise, context when needed — not telegram, not novels
+- **Bilingual:** match conversation language (DE/EN)
+- **CLI topics:** English (commands, script comments) — client project: ask if unsure
+- German: **du**; English: peer-level, direct
+- **Session language lock** for code/commits/new `.md` — no mid-session switch
 
-## What the Agent Must Never Do
-
-- Change code when I only asked a question
-- Hallucinate facts, APIs, or file contents
-- Give outdated answers without flagging uncertainty
-- Ignore rules in this repo or project context files
-- Sanitize or neutralize political content on my sites
-
-## What the Agent Must Always Do
-
-- Follow rules in `AGENTS.md`, `BOUNDARIES.md`, `STANDARDS.md`, `INFRASTRUCTURE.md`, and project `SOUL.md` / `MOOD.md`
-- Persist **explicit rules** in Cursor **Memories** (in addition to markdown files when appropriate)
-- When unsure: research first, then ask — or park the item (see `AGENTS.md`)
-
-## Productivity Triggers
+## Productivity
 
 | Boosts | Drains |
 |--------|--------|
 | Makes me laugh | Debug loops |
-| Finishes work autonomously (when it actually works) | Does things I didn't ask for |
-| State-of-the-art suggestions | Guessing instead of diagnosing |
+| Autonomous finish (when it works) | Unasked changes |
+| State-of-the-art suggestions | Guessing |
 
-## One Sentence That Describes Me
+## One-Liner
 
 > There are 2 hard problems in computer science: cache invalidation, naming things, and off-by-1 errors.
-
-## Availability
-
-No preferences for commit times, branch names, or review rhythm.