diff --git a/.env.example b/.env.example
index fdc36f2..fe0f0c8 100644
--- a/.env.example
+++ b/.env.example
@@ -31,9 +31,6 @@ VOCALIZE_HOST=127.0.0.1
 VOCALIZE_PORT=8080
 ORCHESTRATOR_LISTEN_PORT=8080
 
-# X-Invite-Token shared secret. Required when VOCALIZE_HOST != 127.0.0.1.
-VOCALIZE_INVITE_TOKEN=
-
 # Public WebSocket base URL. Required when VOCALIZE_HOST != 127.0.0.1.
 VOCALIZE_WS_BASE_URL=
 
@@ -49,6 +46,5 @@ LOG_DIR=logs
 # -------------------------------------------------------------------------
 # Frontend (Next.js — baked into the JS bundle at build time)
 # -------------------------------------------------------------------------
-NEXT_PUBLIC_VOCALIZE_API_BASE_URL=https://vocalize-api.dgpisces.com
-NEXT_PUBLIC_VOCALIZE_WS_BASE_URL=wss://vocalize-api.dgpisces.com
-NEXT_PUBLIC_VOCALIZE_INVITE_TOKEN=
+NEXT_PUBLIC_VOCALIZE_API_BASE_URL=https://api.example.com
+NEXT_PUBLIC_VOCALIZE_WS_BASE_URL=wss://api.example.com
diff --git a/.github/ISSUE_TEMPLATE/bug_report.yml b/.github/ISSUE_TEMPLATE/bug_report.yml
index f8fd229..c10026e 100644
--- a/.github/ISSUE_TEMPLATE/bug_report.yml
+++ b/.github/ISSUE_TEMPLATE/bug_report.yml
@@ -5,9 +5,11 @@ body:
   - type: markdown
     attributes:
       value: |
-        Thank you for reporting a bug! Before submitting, please search existing
-        issues to avoid duplicates. For security vulnerabilities, see
-        [SECURITY.md](../../SECURITY.md) instead of filing a public issue.
+        Thank you for reporting a bug! Before submitting, please search
+        existing issues to avoid duplicates. VocalizeAI is a self-deploy
+        project — every operator runs their own backend on their own
+        infrastructure — so all reports (including security-relevant ones)
+        go through GitHub Issues here.
 
   - type: textarea
     id: what-happened
diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
index c5037c6..4341644 100644
--- a/CONTRIBUTING.md
+++ b/CONTRIBUTING.md
@@ -12,7 +12,10 @@ VocalizeAI follows an out-of-band contribution model (D-16):
 3. This keeps the private repo as the single source of truth and preserves the
    `.planning/` workflow.
 
-If your contribution is security-related, see [SECURITY.md](SECURITY.md) first.
+VocalizeAI is a self-deploy project, so there is no centrally hosted
+instance to defend. Security-relevant fixes go through the same PR flow
+as any other contribution; flag the concern in the PR description so
+reviewers know to prioritise it.
 
 ## Setting up the dev environment
 
@@ -67,7 +70,7 @@ test(<area>): <verb> <noun>
 refactor(<area>): <verb> <noun>
 ```
 
-Examples: `feat(server): add X-Invite-Token gate`, `fix(frontend): handle 401 on session create`.
+Examples: `feat(server): bound task length`, `fix(frontend): handle 401 on session create`.
 
 ## Branches
 
@@ -77,8 +80,11 @@ directly to `main`.
 ## Issue triage / vulnerability reporting
 
 - Ordinary bugs and feature requests: file a GitHub issue.
-- Security vulnerabilities: follow the process in [SECURITY.md](SECURITY.md).
-  Do NOT file public GitHub issues for security topics.
+- Security-relevant findings: file a GitHub issue here. VocalizeAI is a
+  self-deploy project (no centrally hosted instance), so there is no
+  separate private disclosure channel. Each operator is responsible for
+  their own deployment; report findings publicly so every operator can
+  pick up the fix.
 
 ## CI behavior for external PRs
 
@@ -110,9 +116,9 @@ VocalizeAI's CI pipeline has two tiers depending on where your PR originates:
 
 VocalizeAI does not adopt a formal Code of Conduct at this stage. Standard
 professional conduct is expected: be respectful, assume good faith, focus on
-the technical content. Disputes that cannot be resolved in-thread escalate to
-the maintainer via email (see SECURITY.md for the contact channel; for
-non-security disputes use the same address).
+the technical content. Disputes that cannot be resolved in-thread escalate
+by opening a GitHub issue with a clear summary; the maintainer will follow
+up there.
 
 ## License
 
diff --git a/README.md b/README.md
index fec8533..ed2aec1 100644
--- a/README.md
+++ b/README.md
@@ -18,16 +18,6 @@ inquiries, status checks, and more. An OSS mirror is available at
 [github.com/DGPisces/VocalizeAI](https://github.com/DGPisces/VocalizeAI)
 under Apache 2.0.
 
-## Why VocalizeAI is invite-only right now
-
-v1 uses a shared invite token (`X-Invite-Token` header on `POST /api/sessions`)
-distributed out-of-band. This is a long-lived shared secret — no rotation flow
-exists in v1; rotation requires a deploy-time env-var change plus frontend
-rebuild. Per-user authentication is v1.x scope (requirement AUTH-01).
-
-Anyone who holds the token can use the service. To request an invite token,
-email [gaodingyun2@gmail.com](mailto:gaodingyun2@gmail.com).
-
 ## Quick Start
 
 **Prerequisites:** Python 3.11+, Node 20+, git, curl. Optional: `uv` (auto-installed by the script).
@@ -121,9 +111,8 @@ VocalizeAI/
 
 | Variable | Purpose |
 |----------|---------|
-| `VOCALIZE_INVITE_TOKEN` | Shared invite token for `POST /api/sessions`; required when `VOCALIZE_HOST != 127.0.0.1` |
-| `VOCALIZE_WS_BASE_URL` | WebSocket base URL returned to clients (e.g., `wss://vocalize-api.example.com`); required in non-localhost mode to prevent Host-header spoofing |
-| `VOCALIZE_CORS_ORIGINS` | Comma-separated allowed CORS origins; defaults to `https://vocalize.example.com` in non-localhost mode |
+| `VOCALIZE_WS_BASE_URL` | WebSocket base URL returned to clients (e.g., `wss://api.example.com`); required in non-localhost mode to prevent Host-header spoofing |
+| `VOCALIZE_CORS_ORIGINS` | Comma-separated allowed CORS origins; **required** in non-localhost mode (no default) |
 
 See `.env.example` for the full env-var inventory including LLM, GPU service,
 and frontend build-time variables.
@@ -205,8 +194,13 @@ follow code style, and submit contributions. Issue + PR templates live under `.g
 
 ## Security
 
-See [SECURITY.md](SECURITY.md) for the vulnerability reporting channel,
-threat model summary, and emergency rollback procedure.
+VocalizeAI is self-deploy: every operator runs their own backend on
+their own infrastructure, and there is no centrally hosted instance to
+defend. Report any security-relevant finding via GitHub Issues — same
+as any other bug — so every operator can pick up the fix. Self-deploy
+operators are responsible for restricting reachability at the network
+or proxy layer (Cloudflare Access, VPN, reverse-proxy auth, etc.).
+Per-user authentication is v1.x scope (requirement `AUTH-01`).
 
 ## License
 
diff --git a/README.zh-CN.md b/README.zh-CN.md
index 7eb3393..570dcf1 100644
--- a/README.zh-CN.md
+++ b/README.zh-CN.md
@@ -15,15 +15,6 @@ clarification_collector / relay）可处理任何电话任务 —— 餐厅订
 [github.com/DGPisces/VocalizeAI](https://github.com/DGPisces/VocalizeAI)，
 协议 Apache 2.0。
 
-## 为什么 VocalizeAI 当前是邀请制
-
-v1 采用共享邀请令牌机制：`POST /api/sessions` 须携带 `X-Invite-Token` 请求头，
-令牌通过线下渠道发放。这是长效共享密钥 —— v1 没有轮换流程；轮换需要修改部署环境变量
-并重新构建前端。每用户认证属于 v1.x 范畴（需求 AUTH-01）。
-
-持有令牌的人均可使用本服务。如需申请邀请令牌，请发邮件至
-[gaodingyun2@gmail.com](mailto:gaodingyun2@gmail.com)。
-
 ## 快速开始
 
 **前提条件：** Python 3.11+、Node 20+、git、curl。可选：`uv`（安装脚本会自动安装）。
@@ -114,9 +105,8 @@ VocalizeAI/
 
 | 变量 | 用途 |
 |------|------|
-| `VOCALIZE_INVITE_TOKEN` | `POST /api/sessions` 所需的共享邀请令牌；`VOCALIZE_HOST != 127.0.0.1` 时必填 |
-| `VOCALIZE_WS_BASE_URL` | 返回给客户端的 WebSocket 基地址（如 `wss://vocalize-api.example.com`）；非 localhost 模式必填，防止 Host 头欺骗 |
-| `VOCALIZE_CORS_ORIGINS` | 允许的 CORS 来源（逗号分隔）；非 localhost 模式默认为 `https://vocalize.example.com` |
+| `VOCALIZE_WS_BASE_URL` | 返回给客户端的 WebSocket 基地址（如 `wss://api.example.com`）；非 localhost 模式必填,防止 Host 头欺骗 |
+| `VOCALIZE_CORS_ORIGINS` | 允许的 CORS 来源（逗号分隔）;非 localhost 模式**必填**(无默认值) |
 
 完整环境变量清单（含 LLM、GPU 服务、前端构建变量）见 `.env.example`。
 
@@ -195,7 +185,11 @@ Issue 模板和 PR 模板位于 `.github/` 目录。
 
 ## 安全
 
-漏洞上报渠道、威胁模型摘要和紧急回滚流程，见 [SECURITY.md](SECURITY.md)。
+VocalizeAI 是自部署项目 —— 每个运维者在自己的基础设施上跑自己的后端,
+没有"统一托管实例"可被攻击。任何安全相关发现请通过 GitHub Issues 上报,
+与普通 bug 同一通道,这样每个运维者都能拿到修复。自部署运维者负责在网络/
+代理层(Cloudflare Access、VPN、反向代理认证等)限制可达性。每用户认证属于
+v1.x 范畴(需求 `AUTH-01`)。
 
 ## 许可证
 
diff --git a/SECURITY.md b/SECURITY.md
deleted file mode 100644
index 2df4f72..0000000
--- a/SECURITY.md
+++ /dev/null
@@ -1,97 +0,0 @@
-# Security Policy
-
-## Supported versions
-
-Only the most recent release tag receives security fixes. v1.x is
-pre-release development — fixes are applied on the main branch and
-shipped with the next release tag.
-
-## Reporting a vulnerability
-
-**Primary channel: GitHub Private Vulnerability Reporting.** Once enabled on
-the public repository (post-Phase-9 setting), open a private security advisory
-at Security → Advisories → New draft advisory. This is the preferred route — it
-keeps the report private and tracked alongside the codebase.
-
-**Backup channel: email.** Until Private Vulnerability Reporting is enabled, or
-if you cannot use it, email
-[gaodingyun2@gmail.com](mailto:gaodingyun2@gmail.com) with subject prefix
-`[VocalizeAI Security]`.
-
-Do **not** file a public GitHub issue for security topics. Disclosure
-follows responsible-disclosure norms: a private 90-day window by default,
-with earlier coordinated release if an active exploit is confirmed.
-
-Please include:
-- Description of the vulnerability and affected component
-- Steps to reproduce
-- Potential impact assessment
-
-## Threat model summary
-
-### Access posture
-
-v1 uses a shared invite-token model. `POST /api/sessions` requires an
-`X-Invite-Token` header whose value must match `VOCALIZE_INVITE_TOKEN`
-configured server-side. This is an intentional design for v1 (requirement
-`AUTH-DEFER-TO-v1.x`); per-user authentication is v1.x scope (requirement
-`AUTH-01`).
-
-### Network layer
-
-- TLS termination at the Cloudflare edge via Universal SSL.
-- A `cloudflared` tunnel connects the Pi orchestrator to the Cloudflare
-  edge; all external traffic enters through Cloudflare before reaching
-  the backend.
-
-### Backend security controls
-
-- **CORS**: single allowed origin (`https://vocalize.dgpisces.com`) in
-  production; localhost origins preserved in dev mode via
-  `VOCALIZE_HOST` env-conditional config. `allow_methods` restricted to
-  `["GET", "POST", "DELETE"]`.
-- **WS base URL enforcement**: server raises at startup if
-  `VOCALIZE_HOST != "127.0.0.1"` and `VOCALIZE_WS_BASE_URL` is unset,
-  preventing Host-header spoofing.
-- **Task length bound**: `SetTaskRequest.task` has a `max_length=2000`
-  field constraint to limit prompt-injection surface area.
-
-## Known limitations
-
-### Shared invite token in JS bundle
-
-`NEXT_PUBLIC_VOCALIZE_INVITE_TOKEN` is embedded in the Next.js JS bundle
-at build time by design — this is inherent to the shared-token model.
-It is not a vulnerability; the token IS the gate. Anyone who can load the
-page already has the token. Per-user secrets are v1.x scope (AUTH-01).
-
-### No token rotation flow in v1
-
-The invite token is a long-lived shared secret. Rotation requires a
-deploy-time env-var update plus a frontend rebuild. There is no in-band
-rotation API in v1.
-
-### No prompt-injection mitigation
-
-User-supplied task descriptions are passed to the LLM without sanitization.
-This is a known gap flagged for v1.x scope.
-
-## Emergency rollback for leaked secret in public mirror
-
-If a secret (invite token, API key, tunnel ID) is accidentally committed
-to the public repo:
-
-1. **Rotate the leaked secret immediately.**
-   - `VOCALIZE_INVITE_TOKEN`: update Pi `.env`, restart `vocalize` service,
-     rebuild frontend with the new `NEXT_PUBLIC_VOCALIZE_INVITE_TOKEN`.
-   - Other secrets: follow their respective rotation procedures.
-
-2. **Force-push the public `main`** to a re-sanitized commit by re-running
-   the `sync-private-to-public` skill against a clean private tree. This
-   overwrites the public history and removes the leaked commit.
-
-3. **Contact GitHub Support** to purge cached references:
-   https://support.github.com/contact/github-private-information-removal
-
-4. **Audit the blast radius**: determine whether the secret was used by
-   any unauthorized party before rotation.
diff --git a/docs/architecture.md b/docs/architecture.md
index f33b82d..b080669 100644
--- a/docs/architecture.md
+++ b/docs/architecture.md
@@ -28,7 +28,7 @@ the AI audio pipeline and the PSTN call.
 
 ### End-to-End Request Flow (Quick Reference)
 
-1. **User creates a session** — `POST /api/sessions` with `X-Invite-Token` → receives
+1. **User creates a session** — `POST /api/sessions` → receives
    `session_id` + `ws_url`.
 2. **User sets the task** — `POST /api/sessions/{id}/task` with `{"task": "..."}` → Phase
    transitions `draft → task_planning → collecting`.
@@ -374,16 +374,17 @@ See: `src/vocalize/server/frames.py`
 
 ## REST Surface
 
-The REST API is mounted at `/api/sessions`. Authentication is via a shared
-invite token (`X-Invite-Token` header on session creation). In localhost-dev
-mode (`VOCALIZE_HOST=127.0.0.1` and `VOCALIZE_INVITE_TOKEN` unset), the token
-gate is disabled.
+The REST API is mounted at `/api/sessions`. The backend ships no
+request-level authentication in v1; self-deploy operators restrict
+reachability at the network or proxy layer (per-user auth is v1.x
+scope — requirement `AUTH-01`).
 
 See: `src/vocalize/server/sessions.py`, `src/vocalize/server/health.py`
 
 ### `POST /api/sessions`
 
-**Auth:** `X-Invite-Token` header (design constraint D-08; uses `secrets.compare_digest`)
+**Auth:** None at the backend layer in v1 (network/proxy restriction is the
+operator's responsibility).
 
 **Request body** (`CreateSessionRequest`, all optional):
 ```json
@@ -405,7 +406,7 @@ See: `src/vocalize/server/sessions.py`, `src/vocalize/server/health.py`
 }
 ```
 
-See: `src/vocalize/server/sessions.py` — `_check_invite_token`, `CreateSessionRequest`
+See: `src/vocalize/server/sessions.py` — `CreateSessionRequest`
 
 ---
 
@@ -565,17 +566,17 @@ See: `src/vocalize/server/ws.py`, `frontend/lib/audio*`, `frontend/components/Br
 ## Security Posture
 
 The security controls relevant to the architecture are documented here for
-API consumers and security researchers. For the full threat model and disclosure
-channel, see [SECURITY.md](../SECURITY.md).
+API consumers and security researchers. VocalizeAI is a self-deploy
+project (no centrally hosted instance); report security-relevant findings
+via GitHub Issues — same channel as any other bug.
 
-### Invite-Token Gate (D-08)
+### Authentication (D-08, retired)
 
-`POST /api/sessions` requires `X-Invite-Token: <token>` matching the server-side
-`VOCALIZE_INVITE_TOKEN` env var. The comparison uses `secrets.compare_digest` to
-prevent timing attacks. In localhost-dev mode (`VOCALIZE_HOST=127.0.0.1` and
-`VOCALIZE_INVITE_TOKEN` unset), the gate is disabled for development convenience.
-
-See: `src/vocalize/server/sessions.py` — `_check_invite_token`
+The original D-08 shared-invite-token gate has been removed; v1 ships no
+backend-level auth on `POST /api/sessions` or the WebSocket. Self-deploy
+operators are expected to restrict reachability at the network or proxy
+layer (Cloudflare Access, VPN, reverse-proxy auth, etc.). Per-user
+authentication is v1.x scope (requirement `AUTH-01`).
 
 ### Task Length Bound (D-09)
 
@@ -612,4 +613,3 @@ See: `src/vocalize/server/ws.py`, `src/vocalize/server/sessions.py`
 - **[docs/deploy/local.md](docs/deploy/local.md)** — Mac/Linux dev environment setup and env-var reference
 - **[docs/deploy/pi.md](docs/deploy/pi.md)** — End-to-end Pi production deployment runbook
 - **[CONTRIBUTING.md](../CONTRIBUTING.md)** — Contributor flow, code style, commit conventions
-- **[SECURITY.md](../SECURITY.md)** — Vulnerability reporting, threat model, emergency rollback
diff --git a/docs/deploy/local.md b/docs/deploy/local.md
index 3181281..631d5fb 100644
--- a/docs/deploy/local.md
+++ b/docs/deploy/local.md
@@ -78,19 +78,18 @@ $EDITOR .env
 | `VOCALIZE_HOST` | default ok | uvicorn bind host; `127.0.0.1` for local dev, `0.0.0.0` for production |
 | `VOCALIZE_PORT` | default ok | uvicorn bind port; default `8080` (note: dev `main.py` defaults to 8000) |
 | `ORCHESTRATOR_LISTEN_PORT` | default ok | Pi service port; default `8080` (legacy; mirrors `VOCALIZE_PORT`) |
-| `VOCALIZE_INVITE_TOKEN` | required when non-localhost | Shared invite secret for `POST /api/sessions`; **gate is disabled when `VOCALIZE_HOST=127.0.0.1` and this var is unset** — localhost-dev shortcut |
-| `VOCALIZE_WS_BASE_URL` | required when non-localhost | Public WS base URL (e.g. `wss://vocalize-api.example.com`); startup raises if missing in non-localhost mode (D-11) |
+| `VOCALIZE_WS_BASE_URL` | required when non-localhost | Public WS base URL (e.g. `wss://api.example.com`); startup raises if missing in non-localhost mode (D-11) |
 | `VOCALIZE_CORS_ORIGINS` | default ok | Comma-separated allowed CORS origins; auto-picked from VOCALIZE_HOST in dev mode |
 | `DEFAULT_LANGUAGE` | default ok | Session default language; `zh` or `en`; default `zh` |
 | `LOG_DIR` | default ok | Log directory; default `logs` |
 | `NEXT_PUBLIC_VOCALIZE_API_BASE_URL` | yes for frontend | Frontend API base URL baked into the Next.js JS bundle at build time |
 | `NEXT_PUBLIC_VOCALIZE_WS_BASE_URL` | optional | Frontend WS base; derived from `NEXT_PUBLIC_VOCALIZE_API_BASE_URL` if absent |
-| `NEXT_PUBLIC_VOCALIZE_INVITE_TOKEN` | yes for frontend in prod | Invite token baked into Next.js JS bundle; required for the frontend to authenticate session creation |
 
-**Localhost-dev shortcut:** when `VOCALIZE_HOST` is `127.0.0.1` and
-`VOCALIZE_INVITE_TOKEN` is unset, the `X-Invite-Token` gate on `POST /api/sessions`
-is disabled. This means you can `curl http://127.0.0.1:8000/api/sessions` without
-supplying a token — convenient for local development.
+**Backend auth posture:** v1 ships no request-level auth on
+`POST /api/sessions` or the WebSocket. For non-localhost deployments,
+restrict reachability at the network or proxy layer (Cloudflare Access,
+VPN, reverse-proxy auth, etc.). Per-user auth is v1.x scope
+(requirement `AUTH-01`).
 
 **Minimum for local dev (no GPU):**
 ```bash
@@ -156,8 +155,7 @@ Exit code 0 = the development environment is working. The smoke script exercises
 6 round-trips: health check, create session, set task, WS upgrade + send/recv,
 delete session. Total runtime is ~20 seconds.
 
-The smoke script uses `VOCALIZE_API_BASE` (default `http://127.0.0.1:8000`) and
-`VOCALIZE_INVITE_TOKEN` (default empty, gate disabled in localhost-dev mode).
+The smoke script uses `VOCALIZE_API_BASE` (default `http://127.0.0.1:8000`).
 
 ---
 
diff --git a/docs/deploy/pi.md b/docs/deploy/pi.md
index 904d501..1a56e20 100644
--- a/docs/deploy/pi.md
+++ b/docs/deploy/pi.md
@@ -127,14 +127,12 @@ sudo nano /opt/vocalize/.env
 | `VOCALIZE_HOST` | default ok | uvicorn bind host; set to `0.0.0.0` for Pi production |
 | `VOCALIZE_PORT` | default ok | uvicorn bind port; default `8080` |
 | `ORCHESTRATOR_LISTEN_PORT` | default ok | Pi service port; default `8080` (legacy compatibility) |
-| `VOCALIZE_INVITE_TOKEN` | **yes** | Shared invite secret for `POST /api/sessions`; generate with `python3 -c 'import secrets; print(secrets.token_urlsafe(32))'` |
-| `VOCALIZE_WS_BASE_URL` | **yes** | Public WS base URL; e.g. `wss://vocalize-api.<your-domain>` — startup raises if missing in non-localhost mode |
+| `VOCALIZE_WS_BASE_URL` | **yes** | Public WS base URL; e.g. `wss://api.<your-domain>` — startup raises if missing in non-localhost mode |
 | `VOCALIZE_CORS_ORIGINS` | default ok | Comma-separated allowed CORS origins; default auto-picked from VOCALIZE_HOST |
 | `DEFAULT_LANGUAGE` | default ok | `zh` or `en`; default `zh` |
 | `LOG_DIR` | default ok | Log directory; default `logs` |
 | `NEXT_PUBLIC_VOCALIZE_API_BASE_URL` | yes (for frontend) | Frontend API base URL; baked into JS bundle at build time |
 | `NEXT_PUBLIC_VOCALIZE_WS_BASE_URL` | optional | Frontend WS base; derived from API base if absent |
-| `NEXT_PUBLIC_VOCALIZE_INVITE_TOKEN` | yes (for frontend in prod) | Shared invite token baked into JS bundle |
 
 **Example production `.env` (use your own values for all `<...>` placeholders):**
 
@@ -147,9 +145,8 @@ SENSEVOICE_WS_PORT=8000
 COSYVOICE_WS_PORT=8001
 VOCALIZE_HOST=0.0.0.0
 VOCALIZE_PORT=8080
-VOCALIZE_INVITE_TOKEN=<random-base64-32>
-VOCALIZE_WS_BASE_URL=wss://vocalize-api.<your-domain>
-VOCALIZE_CORS_ORIGINS=https://vocalize.<your-domain>
+VOCALIZE_WS_BASE_URL=wss://api.<your-domain>
+VOCALIZE_CORS_ORIGINS=https://<your-domain>
 ```
 
 ---
@@ -248,9 +245,7 @@ VOCALIZE_API_BASE=http://127.0.0.1:8080 bash scripts/smoke.sh
 # Exit 0 = working deployment
 
 # Smoke test against the public Cloudflare Tunnel URL:
-VOCALIZE_API_BASE=https://vocalize-api.<your-domain> \
-  VOCALIZE_INVITE_TOKEN=<your-token> \
-  bash scripts/smoke.sh
+VOCALIZE_API_BASE=https://api.<your-domain> bash scripts/smoke.sh
 ```
 
 The smoke script exercises 6 round-trips: `GET /health`, `POST /api/sessions`,
diff --git a/docs/release/24h-stability-evidence.md b/docs/release/24h-stability-evidence.md
index bb893e6..4d7a34d 100644
--- a/docs/release/24h-stability-evidence.md
+++ b/docs/release/24h-stability-evidence.md
@@ -40,7 +40,7 @@ closes this gap:
 | Operator | DGPisces |
 | Date / time (start) | 2026-05-17T19:56:20Z |
 | Date / time (end) | 2026-05-18T19:56:20Z |
-| Backend URL | `http://localhost:8080` (Pi loopback; same orchestrator that `https://vocalize-api.dgpisces.com` fronts via Cloudflare Tunnel) |
+| Backend URL | `http://localhost:8080` (Pi loopback; same orchestrator that `https://api.example.com` fronts via Cloudflare Tunnel) |
 | Browser and version | N/A — driver-only run; real-audio interleave intentionally skipped (see Real-Audio Interleave Log below) |
 | STT service health | Pass (`/health` reported `gpu_reachable: true` throughout) |
 | TTS service health | Pass (CosyVoice docker container up 2 days, healthy) |
@@ -87,7 +87,6 @@ transient failures acceptable).
 Driver command used:
 ```bash
 export VOCALIZE_API_BASE=http://localhost:8080
-export VOCALIZE_INVITE_TOKEN=<token>
 python scripts/stability-24h-driver.py \
   --duration-minutes 1440 \
   --scenario balance_inquiry_en_query \
@@ -95,7 +94,7 @@ python scripts/stability-24h-driver.py \
   --evidence-out docs/release/24h-stability-evidence-runs/2026-05-17T195620Z-pi-loopback.md
 ```
 
-Driver invoked with API base = `http://localhost:8080` (Pi loopback) rather than `https://vocalize-api.dgpisces.com` because the driver ran on the Pi itself per operator preference (full automation, no Attu host). The Cloudflare Tunnel ingress for `vocalize.dgpisces.com` was verified live separately during DEPLOY-03 deploy (curl → HTTP/2 502 + Universal SSL `*.dgpisces.com` cert).
+Driver invoked with API base = `http://localhost:8080` (Pi loopback) rather than `https://api.example.com` because the driver ran on the Pi itself per operator preference (full automation, no Attu host). The Cloudflare Tunnel ingress for `vocalize.example.com` was verified live separately during DEPLOY-03 deploy (curl → HTTP/2 502 + Universal SSL `*.example.com` cert).
 
 ---
 
diff --git a/frontend/app/components.css b/frontend/app/components.css
index 62d9a41..1ea7733 100644
--- a/frontend/app/components.css
+++ b/frontend/app/components.css
@@ -6,7 +6,7 @@
  * except where a value is component-shape (e.g. timer-bar height) and
  * expressed in design-system spacing/timing units.
  *
- * Visual character: dgpisces design system (Apple-style; SF font stack;
+ * Visual character: shared design system (Apple-style; SF font stack;
  * 12px radius cards, 8px sm radius, pill radius for chips). Bubbles and
  * speaker-attributed surfaces follow conversation-app conventions while
  * keeping a refined, minimal tone consistent with the existing B2 chrome.
diff --git a/frontend/app/design-tokens.css b/frontend/app/design-tokens.css
index 28de729..76ab096 100644
--- a/frontend/app/design-tokens.css
+++ b/frontend/app/design-tokens.css
@@ -1,5 +1,5 @@
 /*
- * dgpisces design system — token primitives (web platform).
+ * shared design system — token primitives (web platform).
  *
  * Source: ~/Claude Code/Web Dev/design-system/platforms/web/_tokens.css
  * (mirror of tokens/tokens.json v1.3.0). Vendored here because the design
diff --git a/frontend/app/globals.css b/frontend/app/globals.css
index 1316f72..268b185 100644
--- a/frontend/app/globals.css
+++ b/frontend/app/globals.css
@@ -1,7 +1,7 @@
 /*
  * globals.css — base resets + B2 (browser audio bridge) chrome.
  *
- * Token primitives live in design-tokens.css (vendored from the dgpisces
+ * Token primitives live in design-tokens.css (vendored from the shared
  * design system). Component classes for B3a phases live in components.css.
  * Keep this file scoped to: design-token import, base resets, body chrome,
  * skip-link, and the B2 browser-audio shell (.app-shell / .page-frame /
diff --git a/frontend/lib/api.ts b/frontend/lib/api.ts
index 42d5183..8537e01 100644
--- a/frontend/lib/api.ts
+++ b/frontend/lib/api.ts
@@ -53,34 +53,10 @@ function apiBaseUrl(): string {
   return value.replace(/\/$/, "");
 }
 
-/**
- * Returns the invite token from the build-time env var, or null when unset.
- *
- * Returning null (not throwing) is intentional: localhost-dev must work without
- * any token configured.
- *
- * WARNING: NEXT_PUBLIC_ vars are embedded in the client-side JS bundle and are
- * therefore recoverable by any user who can load the page. This is an accepted
- * v1 limitation — the shared token provides friction, not true access control.
- * Per-user auth (AUTH-01, v1.x) will replace this gate entirely.
- *
- * For stronger protection before AUTH-01 lands, consider routing the session
- * creation call through a Next.js API route that reads the token from a
- * server-side env var (no NEXT_PUBLIC_ prefix) so it never reaches the bundle.
- */
-function inviteToken(): string | null {
-  const value = process.env.NEXT_PUBLIC_VOCALIZE_INVITE_TOKEN;
-  return value && value.length > 0 ? value : null;
-}
-
 export async function createSession(
   body: CreateSessionRequest = {},
 ): Promise<SessionResponse> {
   const headers: Record<string, string> = { "Content-Type": "application/json" };
-  const token = inviteToken();
-  if (token !== null) {
-    headers["X-Invite-Token"] = token;
-  }
   const res = await fetch(`${apiBaseUrl()}/api/sessions`, {
     method: "POST",
     cache: "no-store",
@@ -88,7 +64,6 @@ export async function createSession(
     body: JSON.stringify(body),
   });
   if (!res.ok) {
-    // 401 = invite-required per D-08; bubble to caller for user-facing "invite required" message
     throw new Error(`createSession failed: ${res.status}`);
   }
   return (await res.json()) as SessionResponse;
diff --git a/frontend/scripts/generate-og.py b/frontend/scripts/generate-og.py
index 55cd608..ed5ba04 100644
--- a/frontend/scripts/generate-og.py
+++ b/frontend/scripts/generate-og.py
@@ -5,7 +5,7 @@
   frontend/public/og/og-zh.png  (1200×630, Chinese)
   frontend/public/og/og-en.png  (1200×630, English)
 
-Design system: dgpisces Apple-style tokens
+Design system: Apple-style design tokens
   - Accent:  #007aff (blue) / #5e5ce6 (purple)
   - BG:      linear gradient from #0a0a0f → #1a1a2e
   - Text:    #f5f5f7 (primary), #98989d (soft)
@@ -312,7 +312,7 @@ def generate(lang: str, tagline: str, out_path: str):
         font_url = ImageFont.truetype(FONT_MONO, 15)
     except Exception:
         font_url = ImageFont.load_default()
-    draw.text((tag_x, strip_y), "vocalize.dgpisces.com", font=font_url, fill=TEXT_SOFT)
+    draw.text((tag_x, strip_y), "vocalize.example.com", font=font_url, fill=TEXT_SOFT)
 
     # 10. Decorative divider line (vertical, separating text from illustration)
     div_x = int(W * 0.60)
diff --git a/infra/pi-orchestrator/.env.template b/infra/pi-orchestrator/.env.template
index 6ca1163..dcf758e 100644
--- a/infra/pi-orchestrator/.env.template
+++ b/infra/pi-orchestrator/.env.template
@@ -8,15 +8,10 @@ VOCALIZE_PORT=8080
 
 # Required for non-localhost deployments — see D-11.
 # If unset on non-localhost host, create_app() raises RuntimeError and systemd will restart-loop.
-VOCALIZE_WS_BASE_URL=wss://vocalize-api.dgpisces.com
-
-# Required for the X-Invite-Token gate on POST /api/sessions — see D-08.
-# Generate with: python3 -c 'import secrets; print(secrets.token_urlsafe(32))'
-# MUST be set; an empty value disables the gate, exposing session creation publicly.
-VOCALIZE_INVITE_TOKEN=
+VOCALIZE_WS_BASE_URL=wss://api.example.com
 
 # Production CORS — see D-10. Pin to single origin in prod.
-VOCALIZE_CORS_ORIGINS=https://vocalize.dgpisces.com
+VOCALIZE_CORS_ORIGINS=https://vocalize.example.com
 
 # Test-only: enables merchant_text_inject frame for AI-merchant scenarios.
 # DO NOT set this outside the 24h evidence run.
diff --git a/infra/pi-orchestrator/cloudflared-config.yml b/infra/pi-orchestrator/cloudflared-config.yml
index 8982b47..92f4bbc 100644
--- a/infra/pi-orchestrator/cloudflared-config.yml
+++ b/infra/pi-orchestrator/cloudflared-config.yml
@@ -7,16 +7,16 @@
 #
 # This file documents the intended ingress shape so reviewers can see what the
 # tunnel exposes; the actual ingress lives in the Cloudflare dashboard
-# (Zero Trust → Networks → Tunnels → dgpisces-server1 → Public Hostname) /
+# (Zero Trust → Networks → Tunnels → your-tunnel-name → Public Hostname) /
 # Cloudflare API tunnel config. The dashboard / API is the source of truth —
 # update this YAML whenever the live ingress changes so reviewers of the OSS
 # mirror see the current shape.
 #
-# Tunnel: dgpisces-server1 (<TUNNEL_ID_PLACEHOLDER; configured via cloudflared service install token>)
+# Tunnel: your-tunnel-name (<TUNNEL_ID_PLACEHOLDER; configured via cloudflared service install token>)
 
 ingress:
-  - hostname: vocalize-api.dgpisces.com
+  - hostname: api.example.com
     service: http://localhost:8080
-  - hostname: vocalize.dgpisces.com
+  - hostname: vocalize.example.com
     service: http://localhost:3000
   - service: http_status:404
diff --git a/infra/pi-orchestrator/setup.sh b/infra/pi-orchestrator/setup.sh
index b23d110..99146cd 100755
--- a/infra/pi-orchestrator/setup.sh
+++ b/infra/pi-orchestrator/setup.sh
@@ -88,7 +88,7 @@ echo "     ssh ${PI_USER}@${PI_HOST} 'nano ${PI_PROJECT_DIR}/.env'"
 echo ""
 echo "2. Install cloudflared service on Pi using a tunnel token (token-based auth):"
 echo "   a. Get the connector token from the Cloudflare dashboard:"
-echo "        Zero Trust → Networks → Tunnels → dgpisces-server1 → Configure"
+echo "        Zero Trust → Networks → Tunnels → your-tunnel-name → Configure"
 echo "        → Install and run a connector → copy the long token string"
 echo "   b. SSH to Pi and install the service with that token:"
 echo "        ssh ${PI_USER}@${PI_HOST} 'sudo cloudflared service install <TUNNEL_TOKEN>'"
@@ -97,13 +97,13 @@ echo "      config.yml is needed. (If you ever want to inspect intended ingress,
 echo "      infra/pi-orchestrator/cloudflared-config.yml; that file is reference-only.)"
 echo ""
 echo "3. Make sure Public Hostname routing is configured in the dashboard:"
-echo "        vocalize-api.dgpisces.com → http://localhost:8080"
+echo "        api.example.com → http://localhost:8080"
 echo ""
 echo "4. Start services:"
 echo "     ssh ${PI_USER}@${PI_HOST} 'sudo systemctl start vocalize cloudflared'"
 echo ""
 echo "5. Verify both services are running and the public URL responds:"
 echo "     ssh ${PI_USER}@${PI_HOST} 'sudo systemctl status vocalize cloudflared --no-pager'"
-echo "     curl -fsS https://vocalize-api.dgpisces.com/health"
+echo "     curl -fsS https://api.example.com/health"
 echo ""
 echo "After completing manual steps, use deploy.sh for subsequent deploys."
diff --git a/install/pi-install.sh b/install/pi-install.sh
index 07b2686..511c54b 100755
--- a/install/pi-install.sh
+++ b/install/pi-install.sh
@@ -197,7 +197,7 @@ step6_systemd_unit() {
         run_or_dry "copy .env.template -> ${INSTALL_DIR}/.env" \
             sudo cp "${ENV_SRC}" "${ENV_DST}"
         echo "  Created ${ENV_DST} from template — edit it before starting the service."
-        echo "  At minimum, set: OPENAI_API_KEY, VOCALIZE_INVITE_TOKEN, VOCALIZE_WS_BASE_URL, GPU_HOST"
+        echo "  At minimum, set: OPENAI_API_KEY, VOCALIZE_WS_BASE_URL, VOCALIZE_CORS_ORIGINS, GPU_HOST"
     else
         echo "  ${ENV_DST} already exists — preserving."
     fi
diff --git a/install/public-allowlist.md b/install/public-allowlist.md
index 8664e33..90eae26 100644
--- a/install/public-allowlist.md
+++ b/install/public-allowlist.md
@@ -15,7 +15,6 @@ Paths are anchored at repo root. Used by `scripts/build-public-filelist.py`.
 - CHANGELOG.md
 - CODEOWNERS
 - LICENSE
-- SECURITY.md
 - CONTRIBUTING.md
 - pyproject.toml
 - .env.example
diff --git a/pyproject.toml b/pyproject.toml
index 20d88b7..77394c5 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -6,7 +6,7 @@ build-backend = "setuptools.build_meta"
 name = "vocalize-ai"
 version = "0.1.0"
 description = "Bilingual (zh/en) universal phone-task AI agent"
-authors = [{name = "DGPisces", email = "gaodingyun2@gmail.com"}]
+authors = [{name = "DGPisces"}]
 license = "Apache-2.0"
 readme = "README.md"
 requires-python = ">=3.11"
diff --git a/scripts/smoke.sh b/scripts/smoke.sh
index fa8c4a3..e5cf46e 100755
--- a/scripts/smoke.sh
+++ b/scripts/smoke.sh
@@ -15,14 +15,13 @@ IFS=$'\n\t'
 # Usage:
 #   bash scripts/smoke.sh
 #   VOCALIZE_API_BASE=http://127.0.0.1:8080 bash scripts/smoke.sh
-#   VOCALIZE_INVITE_TOKEN=<token> VOCALIZE_API_BASE=https://api.example.com bash scripts/smoke.sh
+#   VOCALIZE_API_BASE=https://api.example.com bash scripts/smoke.sh
 #
 # Exits 0 on all-pass; exits 1 with descriptive message on any failure.
 # Total runtime budget: ~20 seconds (WS step has a 10s timeout).
 # ---------------------------------------------------------------------------
 
 BASE="${VOCALIZE_API_BASE:-http://127.0.0.1:8000}"
-TOKEN="${VOCALIZE_INVITE_TOKEN:-}"
 
 SMOKE_START=$(date +%s)
 SID=""
@@ -81,10 +80,9 @@ echo ""
 echo "[2/5] POST ${BASE}/api/sessions..."
 
 SESSIONS_RESP=$(curl -fsS --max-time 10 -X POST "${BASE}/api/sessions" \
-    -H "X-Invite-Token: ${TOKEN}" \
     -H "Content-Type: application/json" \
     -d '{"default_lang":"zh","auto_translate_merchant":true}' 2>/dev/null) \
-    || fail "[2/5]" "curl failed — is the backend running and invite token correct?"
+    || fail "[2/5]" "curl failed — is the backend running?"
 
 SID=$(echo "$SESSIONS_RESP" | jq -r '.session_id' 2>/dev/null) \
     || fail "[2/5]" "response is not valid JSON: ${SESSIONS_RESP}"
diff --git a/scripts/stability-24h-driver.py b/scripts/stability-24h-driver.py
index 1692c53..a0dec12 100644
--- a/scripts/stability-24h-driver.py
+++ b/scripts/stability-24h-driver.py
@@ -7,20 +7,16 @@
 Usage (run from a remote runner host):
     ssh <your-runner-host>
     cd /path/to/VocalizeAI
-    export VOCALIZE_API_BASE=https://vocalize-api.dgpisces.com
-    export VOCALIZE_INVITE_TOKEN=<token>
+    export VOCALIZE_API_BASE=https://api.example.com
     python scripts/stability-24h-driver.py --duration-minutes 1440 \\
         --scenario balance_inquiry_en_query --seed direct
 
 Environment variables:
     VOCALIZE_API_BASE     Base URL for the Pi orchestrator REST API.
-                          Default: https://vocalize-api.dgpisces.com
-    VOCALIZE_INVITE_TOKEN Required when API base is non-localhost.
-                          Set to match VOCALIZE_INVITE_TOKEN on the Pi.
+                          Default: https://api.example.com
 
 Prerequisites on Pi (MUST be set before the 24h run):
     VOCALIZE_ENABLE_TEST_FRAMES=1  in /opt/vocalize/.env (or your VOCALIZE_HOME path)
-    VOCALIZE_INVITE_TOKEN=<token>  in /opt/vocalize/.env (or your VOCALIZE_HOME path)
 
 After the run, paste the output file into
 docs/release/24h-stability-evidence.md for the release record.
@@ -53,7 +49,7 @@
 DEFAULT_DURATION_MIN = 24 * 60  # 24 hours
 DEFAULT_SCENARIO_ID = "balance_inquiry_en_query"
 DEFAULT_SEED_ID = "direct"
-DEFAULT_API_BASE = "https://vocalize-api.dgpisces.com"
+DEFAULT_API_BASE = "https://api.example.com"
 
 # Path to scenarios.yaml (sibling of tests/integration/ai_merchant.py)
 _SCENARIOS_YAML = (
@@ -118,7 +114,6 @@ def _get_merchant_lang(scenario_id: str) -> str:
 # ---------------------------------------------------------------------------
 async def one_cycle(
     api_base: str,
-    invite_token: str | None,
     scenario_id: str,
     seed_id: str,
 ) -> dict[str, Any]:
@@ -131,8 +126,6 @@ async def one_cycle(
     """
     cycle_start = time.time()
     headers: dict[str, str] = {}
-    if invite_token:
-        headers["X-Invite-Token"] = invite_token
 
     user_lang = _get_user_lang(scenario_id)
     merchant_lang = _get_merchant_lang(scenario_id)
@@ -263,14 +256,9 @@ async def one_cycle(
 # ---------------------------------------------------------------------------
 # Metrics snapshot
 # ---------------------------------------------------------------------------
-async def fetch_metrics_snapshot(
-    api_base: str,
-    invite_token: str | None = None,
-) -> str:
+async def fetch_metrics_snapshot(api_base: str) -> str:
     """GET {api_base}/metrics and return the Prometheus exposition text."""
     headers: dict[str, str] = {}
-    if invite_token:
-        headers["X-Invite-Token"] = invite_token
     try:
         async with httpx.AsyncClient(timeout=15.0) as client:
             resp = await client.get(f"{api_base}/metrics", headers=headers)
@@ -370,16 +358,6 @@ def _write_evidence(
 # ---------------------------------------------------------------------------
 async def main(args: argparse.Namespace) -> None:
     api_base = args.api_base
-    invite_token = args.invite_token
-
-    # Require token for non-localhost
-    is_localhost = api_base.startswith("http://127.") or api_base.startswith("http://localhost")
-    if not is_localhost and not invite_token:
-        log.error(
-            "VOCALIZE_INVITE_TOKEN is required when VOCALIZE_API_BASE is not localhost.\n"
-            "Set it in your environment before running the driver."
-        )
-        sys.exit(1)
 
     log.info(
         "stability-24h-driver starting: api_base=%s scenario=%s seed=%s duration_min=%d",
@@ -393,7 +371,7 @@ async def main(args: argparse.Namespace) -> None:
     metrics_snapshots: list[tuple[float, str]] = []
 
     # Initial metrics snapshot
-    snap = await fetch_metrics_snapshot(api_base, invite_token)
+    snap = await fetch_metrics_snapshot(api_base)
     metrics_snapshots.append((time.time(), snap))
     log.info("initial /metrics snapshot taken")
 
@@ -403,7 +381,7 @@ async def main(args: argparse.Namespace) -> None:
         cycle_start = time.time()
         log.info("--- cycle %d start (remaining: %.0f min) ---", cycle_num, (end - cycle_start) / 60)
 
-        result = await one_cycle(api_base, invite_token, args.scenario, args.seed)
+        result = await one_cycle(api_base, args.scenario, args.seed)
         session_log.append(result)
         status = "OK" if result["ok"] else f"FAIL: {result.get('error', '')}"
         log.info("cycle %d result: %s (%.1fs)", cycle_num, status, result.get("elapsed_s", 0))
@@ -411,7 +389,7 @@ async def main(args: argparse.Namespace) -> None:
         # Periodic metrics snapshot
         last_snap_t = metrics_snapshots[-1][0]
         if (cycle_start - last_snap_t) >= args.metrics_snapshot_interval_hours * 3600:
-            snap = await fetch_metrics_snapshot(api_base, invite_token)
+            snap = await fetch_metrics_snapshot(api_base)
             metrics_snapshots.append((time.time(), snap))
             log.info("metrics snapshot taken at cycle %d", cycle_num)
 
@@ -423,7 +401,7 @@ async def main(args: argparse.Namespace) -> None:
             await asyncio.sleep(min(sleep_s, remaining))
 
     # Final metrics snapshot
-    snap = await fetch_metrics_snapshot(api_base, invite_token)
+    snap = await fetch_metrics_snapshot(api_base)
     metrics_snapshots.append((time.time(), snap))
     log.info("final /metrics snapshot taken; %d cycles completed", cycle_num)
 
@@ -484,7 +462,6 @@ def _build_arg_parser() -> argparse.ArgumentParser:
 
     # Resolve env vars
     args.api_base = os.environ.get("VOCALIZE_API_BASE", DEFAULT_API_BASE).rstrip("/")
-    args.invite_token = os.environ.get("VOCALIZE_INVITE_TOKEN") or None
 
     # Default evidence output path
     if args.evidence_out is None:
diff --git a/src/vocalize/config.py b/src/vocalize/config.py
index ae75d7e..efccbed 100644
--- a/src/vocalize/config.py
+++ b/src/vocalize/config.py
@@ -53,9 +53,6 @@ class Config:
     # 日志配置
     log_dir: str = "logs"
 
-    # Phase 4 D-08: X-Invite-Token 共享密钥；None = 本地开发模式，跳过邀请验证
-    invite_token: str | None = None
-
     @classmethod
     def from_env(cls) -> "Config":
         """从环境变量和 .env 文件加载配置。"""
@@ -74,7 +71,6 @@ def from_env(cls) -> "Config":
             ),
             default_language=os.getenv("DEFAULT_LANGUAGE", cls.default_language),
             log_dir=os.getenv("LOG_DIR", cls.log_dir),
-            invite_token=os.getenv("VOCALIZE_INVITE_TOKEN") or None,
         )
 
     def validate_for_phase(
diff --git a/src/vocalize/server/__init__.py b/src/vocalize/server/__init__.py
index 4396792..8e0de69 100644
--- a/src/vocalize/server/__init__.py
+++ b/src/vocalize/server/__init__.py
@@ -47,7 +47,7 @@ def _default_user_pipeline_factory(transport):
     )
 
 
-_DEFAULT_PROD_ORIGINS = ["https://vocalize.dgpisces.com"]
+_DEFAULT_PROD_ORIGINS: list[str] = []
 _DEFAULT_DEV_ORIGINS = ["http://localhost:3000", "http://127.0.0.1:3000"]
 
 
@@ -58,14 +58,12 @@ def create_app() -> FastAPI:
         VOCALIZE_HOST / VOCALIZE_PORT — where uvicorn binds (handled by main()).
         VOCALIZE_CORS_ORIGINS — comma-separated allowed origins; defaults to
             dev origins when VOCALIZE_HOST is 127.0.0.1/localhost, else
-            the single production origin https://vocalize.dgpisces.com (D-10).
+            empty (operator MUST set this env var in non-localhost mode — see D-10).
         VOCALIZE_WS_BASE_URL — REQUIRED when VOCALIZE_HOST is not localhost.
             The public WS prefix echoed back from POST /api/sessions.
             Raises RuntimeError at startup when absent in non-localhost mode
             (closes Host-header spoofing vector D-11 — see CONCERNS.md).
             In localhost-dev mode the WS URL is derived from the request base_url.
-        VOCALIZE_INVITE_TOKEN — shared invite secret for POST /api/sessions;
-            gate disabled in localhost-dev mode when unset (D-08).
         GPU_HOST / SENSEVOICE_WS_PORT / COSYVOICE_WS_PORT — GPU service targets.
     """
     app = FastAPI(title="VocalizeAI", version="0.1.0")
@@ -99,7 +97,7 @@ def create_app() -> FastAPI:
         allow_origins=cors_origins,
         allow_credentials=False,
         allow_methods=["GET", "POST", "DELETE"],  # D-10: explicit list; no wildcards
-        allow_headers=["Content-Type", "X-Invite-Token"],  # explicit; covers auth header
+        allow_headers=["Content-Type"],  # explicit; no wildcards
     )
 
     registry = SessionRegistry()
@@ -126,7 +124,7 @@ async def _refresh_runtime_gauges_on_metrics_scrape(
         raise RuntimeError(
             "VOCALIZE_WS_BASE_URL is required when VOCALIZE_HOST is not localhost "
             "(closes Host-header spoofing vector — see CONCERNS.md). "
-            "Example: wss://vocalize-api.dgpisces.com"
+            "Example: wss://api.example.com"
         )
 
     register_session_routes(app, registry=registry)
diff --git a/src/vocalize/server/sessions.py b/src/vocalize/server/sessions.py
index 28f6704..848386c 100644
--- a/src/vocalize/server/sessions.py
+++ b/src/vocalize/server/sessions.py
@@ -7,10 +7,9 @@
 
 import logging
 import os
-import secrets
 from typing import Literal
 
-from fastapi import Depends, FastAPI, Header, HTTPException, Request, status
+from fastapi import FastAPI, HTTPException, Request
 from pydantic import BaseModel, Field
 
 from vocalize.server.review import register_review_routes
@@ -52,45 +51,6 @@ class SetTaskResponse(BaseModel):
     ok: bool = True
 
 
-def _check_invite_token(
-    x_invite_token: str | None = Header(default=None, alias="X-Invite-Token"),
-) -> None:
-    """Verify the shared invite secret on session creation (D-08).
-
-    Gate behaviour:
-    - If VOCALIZE_INVITE_TOKEN is not configured (localhost-dev mode), the
-      check is skipped so local development keeps working without any env setup.
-    - In production (non-localhost host), the token is required and must
-      match via constant-time comparison to avoid timing oracles (T-04c-02).
-
-    # TODO(v1.x AUTH-01): no rotation; this is a long-lived shared secret.
-    #   Per-user auth replaces this gate in v1.x.
-    """
-    from vocalize.config import get_config
-
-    expected = get_config().invite_token
-    if expected is None:
-        return  # localhost-dev mode: gate disabled
-    try:
-        match = secrets.compare_digest(x_invite_token or "", expected)
-    except (TypeError, ValueError):
-        # secrets.compare_digest raises TypeError when either string contains
-        # non-ASCII characters. This is operator misconfiguration (e.g. a
-        # Unicode passphrase). Log once and return 401 — do not propagate as
-        # a 500 which would increment vocalize_error_log_total toward the
-        # D-05 budget (T-04c-02).
-        log.warning(
-            "VOCALIZE_INVITE_TOKEN contains non-ASCII characters; "
-            "compare_digest raised TypeError — returning 401"
-        )
-        match = False
-    if not match:
-        raise HTTPException(
-            status_code=status.HTTP_401_UNAUTHORIZED,
-            detail="invalid or missing X-Invite-Token",
-        )
-
-
 def register_session_routes(
     app: FastAPI,
     *,
@@ -108,7 +68,6 @@ def register_session_routes(
     async def create_session(
         request: Request,
         payload: CreateSessionRequest | None = None,
-        _gate: None = Depends(_check_invite_token),
     ) -> CreateSessionResponse:
         registry.sweep_stale(max_age_s=1800)
         payload = payload or CreateSessionRequest()
diff --git a/tests/test_app_startup_guards.py b/tests/test_app_startup_guards.py
index 7d377b0..9a7cfe6 100644
--- a/tests/test_app_startup_guards.py
+++ b/tests/test_app_startup_guards.py
@@ -29,7 +29,6 @@ def _make_client(monkeypatch, *, host: str, ws_base_url: str | None = None) -> T
         monkeypatch.setenv("VOCALIZE_WS_BASE_URL", ws_base_url)
     else:
         monkeypatch.delenv("VOCALIZE_WS_BASE_URL", raising=False)
-    monkeypatch.delenv("VOCALIZE_INVITE_TOKEN", raising=False)
     monkeypatch.delenv("VOCALIZE_CORS_ORIGINS", raising=False)
     reset_config()
     return TestClient(create_app())
@@ -56,7 +55,6 @@ def test_localhost_without_ws_base_url_succeeds(monkeypatch):
 def test_non_localhost_without_ws_base_url_raises(monkeypatch):
     monkeypatch.setenv("VOCALIZE_HOST", "0.0.0.0")
     monkeypatch.delenv("VOCALIZE_WS_BASE_URL", raising=False)
-    monkeypatch.delenv("VOCALIZE_INVITE_TOKEN", raising=False)
     monkeypatch.delenv("VOCALIZE_CORS_ORIGINS", raising=False)
     reset_config()
     with pytest.raises(RuntimeError, match="VOCALIZE_WS_BASE_URL is required"):
@@ -71,7 +69,7 @@ def test_non_localhost_with_ws_base_url_succeeds(monkeypatch):
     client = _make_client(
         monkeypatch,
         host="0.0.0.0",
-        ws_base_url="wss://vocalize-api.dgpisces.com",
+        ws_base_url="wss://api.example.com",
     )
     # App started without RuntimeError; 404 here just means session not found.
     resp = client.get("/api/sessions/nonexistent-id")
@@ -89,7 +87,7 @@ def test_cors_localhost_origin_allowed(monkeypatch):
         headers={
             "Origin": "http://localhost:3000",
             "Access-Control-Request-Method": "POST",
-            "Access-Control-Request-Headers": "X-Invite-Token",
+            "Access-Control-Request-Headers": "Content-Type",
         },
     )
     assert resp.status_code == 200
@@ -124,7 +122,7 @@ def test_cors_preflight_returns_explicit_methods_and_headers(monkeypatch):
         headers={
             "Origin": "http://localhost:3000",
             "Access-Control-Request-Method": "POST",
-            "Access-Control-Request-Headers": "Content-Type, X-Invite-Token",
+            "Access-Control-Request-Headers": "Content-Type",
         },
     )
     assert resp.status_code == 200
@@ -135,8 +133,8 @@ def test_cors_preflight_returns_explicit_methods_and_headers(monkeypatch):
     assert "POST" in allow_methods
     assert "DELETE" in allow_methods
     assert "*" not in allow_methods
-    # Headers must include X-Invite-Token
-    assert "X-Invite-Token" in allow_headers
+    # Headers must include Content-Type
+    assert "Content-Type" in allow_headers
 
 
 # ---------------------------------------------------------------------------
@@ -147,7 +145,6 @@ def test_cors_wildcard_origins_raises_at_startup(monkeypatch):
     monkeypatch.setenv("VOCALIZE_HOST", "127.0.0.1")
     monkeypatch.setenv("VOCALIZE_CORS_ORIGINS", "*")
     monkeypatch.delenv("VOCALIZE_WS_BASE_URL", raising=False)
-    monkeypatch.delenv("VOCALIZE_INVITE_TOKEN", raising=False)
     reset_config()
     with pytest.raises(RuntimeError, match="VOCALIZE_CORS_ORIGINS must not contain"):
         create_app()
@@ -159,9 +156,8 @@ def test_cors_wildcard_origins_raises_at_startup(monkeypatch):
 
 def test_cors_wildcard_mixed_with_real_origins_raises(monkeypatch):
     monkeypatch.setenv("VOCALIZE_HOST", "0.0.0.0")
-    monkeypatch.setenv("VOCALIZE_WS_BASE_URL", "wss://vocalize-api.dgpisces.com")
+    monkeypatch.setenv("VOCALIZE_WS_BASE_URL", "wss://api.example.com")
     monkeypatch.setenv("VOCALIZE_CORS_ORIGINS", "https://example.com,*")
-    monkeypatch.delenv("VOCALIZE_INVITE_TOKEN", raising=False)
     reset_config()
     with pytest.raises(RuntimeError, match="VOCALIZE_CORS_ORIGINS must not contain"):
         create_app()
diff --git a/tests/test_invite_token_gate.py b/tests/test_invite_token_gate.py
deleted file mode 100644
index 9682381..0000000
--- a/tests/test_invite_token_gate.py
+++ /dev/null
@@ -1,160 +0,0 @@
-"""Tests for the X-Invite-Token gate on POST /api/sessions (D-08).
-
-Covers:
-  - Localhost-dev mode (no token configured): gate disabled, POST succeeds.
-  - Production mode (token configured): gate enforces header presence + value.
-  - SetTaskRequest.task length bound (max_length=2000).
-  - Source assertion: _check_invite_token uses secrets.compare_digest.
-"""
-from __future__ import annotations
-
-import inspect
-
-import pytest
-from fastapi.testclient import TestClient
-
-from vocalize.config import reset_config
-from vocalize.server import create_app
-
-
-def _client(monkeypatch) -> TestClient:
-    """Build a TestClient for the full production app.
-
-    VOCALIZE_HOST is forced to 127.0.0.1 so create_app() does not raise the
-    startup RuntimeError for missing VOCALIZE_WS_BASE_URL (Task 2).
-    VOCALIZE_CORS_ORIGINS and VOCALIZE_WS_BASE_URL are cleared to prevent a
-    CI environment with VOCALIZE_CORS_ORIGINS='*' from raising RuntimeError
-    inside create_app() and masking the invite-token test failures.
-    """
-    monkeypatch.setenv("VOCALIZE_HOST", "127.0.0.1")
-    monkeypatch.delenv("VOCALIZE_CORS_ORIGINS", raising=False)
-    monkeypatch.delenv("VOCALIZE_WS_BASE_URL", raising=False)
-    reset_config()
-    return TestClient(create_app())
-
-
-# ---------------------------------------------------------------------------
-# Test 1: No token configured → localhost-dev gate disabled, POST succeeds
-# ---------------------------------------------------------------------------
-
-def test_no_token_configured_allows_session_creation(monkeypatch):
-    monkeypatch.delenv("VOCALIZE_INVITE_TOKEN", raising=False)
-    client = _client(monkeypatch)
-    resp = client.post("/api/sessions")
-    assert resp.status_code == 200
-    assert "session_id" in resp.json()
-
-
-# ---------------------------------------------------------------------------
-# Test 2: Token configured, header missing → 401
-# ---------------------------------------------------------------------------
-
-def test_missing_header_returns_401_when_token_configured(monkeypatch):
-    monkeypatch.setenv("VOCALIZE_INVITE_TOKEN", "abc123")
-    client = _client(monkeypatch)
-    resp = client.post("/api/sessions")
-    assert resp.status_code == 401
-    assert resp.json()["detail"] == "invalid or missing X-Invite-Token"
-
-
-# ---------------------------------------------------------------------------
-# Test 3: Token configured, wrong header value → 401
-# ---------------------------------------------------------------------------
-
-def test_wrong_header_value_returns_401(monkeypatch):
-    monkeypatch.setenv("VOCALIZE_INVITE_TOKEN", "abc123")
-    client = _client(monkeypatch)
-    resp = client.post("/api/sessions", headers={"X-Invite-Token": "wrong-value"})
-    assert resp.status_code == 401
-    assert resp.json()["detail"] == "invalid or missing X-Invite-Token"
-
-
-# ---------------------------------------------------------------------------
-# Test 4: Token configured, correct header value → 200 + session_id
-# ---------------------------------------------------------------------------
-
-def test_correct_header_value_allows_session_creation(monkeypatch):
-    monkeypatch.setenv("VOCALIZE_INVITE_TOKEN", "abc123")
-    client = _client(monkeypatch)
-    resp = client.post("/api/sessions", headers={"X-Invite-Token": "abc123"})
-    assert resp.status_code == 200
-    data = resp.json()
-    assert "session_id" in data
-    assert data["session_id"]
-
-
-# ---------------------------------------------------------------------------
-# Test 5: SetTaskRequest.task > 2000 chars → 422
-# ---------------------------------------------------------------------------
-
-def test_task_length_over_2000_returns_422(monkeypatch):
-    monkeypatch.delenv("VOCALIZE_INVITE_TOKEN", raising=False)
-    client = _client(monkeypatch)
-    # Create a session first
-    sess_resp = client.post("/api/sessions")
-    assert sess_resp.status_code == 200
-    session_id = sess_resp.json()["session_id"]
-
-    oversized_task = "x" * 2001
-    resp = client.post(
-        f"/api/sessions/{session_id}/task",
-        json={"task": oversized_task},
-    )
-    assert resp.status_code == 422
-
-
-# ---------------------------------------------------------------------------
-# Test 6: SetTaskRequest.task == 2000 chars → 200
-# ---------------------------------------------------------------------------
-
-def test_task_length_exactly_2000_is_accepted(monkeypatch):
-    monkeypatch.delenv("VOCALIZE_INVITE_TOKEN", raising=False)
-    client = _client(monkeypatch)
-    sess_resp = client.post("/api/sessions")
-    assert sess_resp.status_code == 200
-    session_id = sess_resp.json()["session_id"]
-
-    max_task = "x" * 2000
-    resp = client.post(
-        f"/api/sessions/{session_id}/task",
-        json={"task": max_task},
-    )
-    assert resp.status_code == 200
-
-
-# ---------------------------------------------------------------------------
-# Test 7: Source assertion — _check_invite_token uses secrets.compare_digest
-# ---------------------------------------------------------------------------
-
-def test_check_invite_token_uses_timing_safe_comparison():
-    from vocalize.server.sessions import _check_invite_token  # noqa: PLC0415
-
-    source = inspect.getsource(_check_invite_token)
-    assert "secrets.compare_digest" in source, (
-        "_check_invite_token must use secrets.compare_digest for timing-safe "
-        "comparison (T-04c-02)"
-    )
-
-
-# ---------------------------------------------------------------------------
-# Test 8: Non-ASCII invite token → 401 (not 500) — regression for WR-04
-# ---------------------------------------------------------------------------
-
-def test_non_ascii_invite_token_returns_401_not_500(monkeypatch):
-    """secrets.compare_digest raises TypeError when VOCALIZE_INVITE_TOKEN contains
-    non-ASCII characters (e.g. a Unicode passphrase set by the operator).
-
-    HTTP headers are ASCII-only, so the client always sends an ASCII (or empty)
-    token value; the non-ASCII value lives in the expected side only.
-    The gate must catch TypeError and return 401, not propagate a 500 that would
-    increment vocalize_error_log_total toward the D-05 error budget (T-04c-02).
-    """
-    monkeypatch.setenv("VOCALIZE_INVITE_TOKEN", "パスワード123")
-    client = _client(monkeypatch)
-    # Send an ASCII token value; the server side holds the non-ASCII expected value.
-    # compare_digest raises TypeError when either argument is non-ASCII.
-    resp = client.post("/api/sessions", headers={"X-Invite-Token": "ascii-token"})
-    assert resp.status_code == 401, (
-        f"Expected 401 for non-ASCII configured token, got {resp.status_code}"
-    )
-    assert resp.json()["detail"] == "invalid or missing X-Invite-Token"