relmio 0.5.0 → 0.7.0

package/CHANGELOG.md

@@ -7,6 +7,56 @@ checks the registry separately after publication.
 
 ## Unreleased
 
+## [0.7.0] - 2026-08-16
+
+### Added
+
+- Add an encrypted in-wizard tester for the experimental Chat Adapter, plus
+  safe sample Chat Adapter and Codex App Server commands for local testing.
+- Add generated hosted guides for getting started, local endpoints, VPS and
+  n8n, troubleshooting, FAQ, security, and reference information.
+
+### Changed
+
+- Synchronize concise root and npm READMEs around product, installation,
+  security, and common-problem overviews that link to hosted guides.
+- Fact-check ChatGPT/Codex token-refresh guidance across documentation: tokens
+  refresh during active use, the official documentation specifies no fixed
+  10-day lifetime, and the provider credential remains distinct from Relmio's
+  rotatable client capability.
+
+### Security
+
+- Limit tester destinations to literal loopback HTTP addresses, retain private
+  keys only in memory for a bounded lifetime, encrypt entered credentials
+  before they cross the browser boundary, require POST after a completed Chat
+  Adapter install, keep sample bearer values out of process arguments, and
+  erase or abort sessions when forgotten, rotated, or shut down.
+
+## [0.6.0] - 2026-08-15
+
+### Added
+
+- Add an experimental loopback-only Codex Chat Adapter for trusted local
+  backends and development servers, with bearer authentication, multi-turn
+  conversation IDs, strict resource bounds, and a small Relmio-specific
+  `POST /chat` contract.
+
+### Changed
+
+- Make Codex device sign-in target-aware so the experimental Relmio `/chat`
+  adapter and native App Server retain isolated, persistent ChatGPT credentials;
+  a Platform API key powers neither target and remains reserved for the generic
+  OpenAI-compatible `/v1` endpoint.
+
+### Security
+
+- Reject browser-origin adapter requests, keep the adapter separate from
+  Platform-key-backed generic OpenAI-compatible `/v1` semantics, explicitly deny
+  model turns access to the private Codex credential store, run chat turns
+  read-only without network access, and preserve loopback-only publication plus
+  credential rotation.
+
 ## [0.5.0] - 2026-08-15
 
 ### Added

package/README.md

@@ -1,433 +1,61 @@
-<div align="center">
-  <img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/brand/relmio-mark.svg" alt="Relmio logo" width="88">
-  <h1>Relmio</h1>
-  <p>Set up private n8n relays, Platform-key OpenAI-compatible local endpoints, and experimental Codex App Server sessions.</p>
-  <p>
-    <a href="https://github.com/Demonbane18/relmio">Full guide</a>
-    &nbsp;·&nbsp;
-    <a href="https://relmio.vercel.app/">Hosted ChatGPT site</a>
-    &nbsp;·&nbsp;
-    <a href="https://github.com/Demonbane18/relmio/issues/new">Report an issue</a>
-    &nbsp;·&nbsp;
-    <a href="https://github.com/Demonbane18/relmio/blob/main/docs/roadmap.md">Roadmap</a>
-  </p>
-  <p>
-    <a href="https://ko-fi.com/paldogies"><img src="https://img.shields.io/badge/Ko--fi-support-ff5e5b.svg?logo=ko-fi&logoColor=white" alt="Support Relmio on Ko-fi"></a>
-  </p>
-</div>
+# Relmio
 
-## See it working first
+This package starts Relmio's local setup wizard for a private n8n sidecar or a
+loopback-only local endpoint. It supports Node.js 22 or newer.
 
-This documented example follows a tested n8n OpenAI credential through a
-published Telegram-triggered workflow. It is a product-operation record, not
-an endorsement, sponsorship, or affiliation by OpenAI, n8n, Hostinger,
-Telegram, or AppBuildersPH.
-
-<figure>
-  <img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/examples/n8n-openai-credential-connected.png" alt="n8n OpenAI credential dialog showing that the connection test succeeded" width="960">
-  <figcaption>The n8n OpenAI credential connection test succeeded; the credential value itself is obscured.</figcaption>
-</figure>
-
-<figure>
-  <img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/examples/gpt-56-model-selector.png" alt="n8n model selector with gpt-5.6-terra selected and an account-specific model list" width="500">
-  <figcaption>This signed-in account's n8n model list includes <code>gpt-5.6-terra</code>; model availability is account-dependent and can change.</figcaption>
-</figure>
-
-<figure>
-  <img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/examples/telegram-n8n-workflow-execution.png" alt="Successful n8n execution of a Telegram-triggered workflow with HTTP Request, Basic LLM Chain, and AI Agent branches" width="960">
-  <figcaption>The published Telegram-triggered n8n workflow completed successfully across its HTTP Request, Basic LLM Chain, and AI Agent branches.</figcaption>
-</figure>
-
-<figure>
-  <img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/examples/telegram-model-results.png" alt="Telegram conversation receiving HTTP Request, Basic LLM Chain, and AI Agent outputs" width="682">
-  <figcaption>Telegram received the HTTP Request, Basic LLM Chain, and AI Agent outputs sent by that workflow.</figcaption>
-</figure>
-
-Relmio is a local browser wizard with separate setup paths. Its existing
-VPS/n8n path installs a private
-[openai-oauth](https://github.com/EvanZhouDev/openai-oauth) Docker sidecar
-beside an existing self-hosted n8n instance. Its local Docker path can install
-either an OpenAI-compatible gateway backed by a Platform API key or the
-official Codex App Server backed by ChatGPT sign-in.
-
-The existing n8n image, Compose file, container, and workflows stay untouched.
-
-A sidecar is a small helper program that runs beside a larger program. Relmio's
-private sidecar adds ChatGPT/Codex sign-in and request translation while your
-existing n8n stays unchanged. This follows the sidecar pattern described in
-[Justin Rice's beginner-friendly overview](https://medium.com/@justinricedev/what-is-a-software-sidecar-8f89feff09f9).
-
-Think of it like a motorcycle gaining a sidecar: together they become a
-tricycle with extra seats. n8n is the motorcycle; Relmio is the sidecar that
-adds the missing capability without changing n8n.
-
-Try the hosted browser demo at
-[relmio.vercel.app](https://relmio.vercel.app/). It is a separate
-request-bound ChatGPT experience; the npm package remains the local wizard for
-the private VPS/n8n sidecar and the separate local Docker endpoint paths.
-
-> **Warning — Hosted chat requires the browser extension.** Install the
-> open-source [Sign in with ChatGPT extension for
-> Chrome](https://chromewebstore.google.com/detail/sign-in-with-chatgpt/odbgboachaefbbbdiffcefhpkekhfcna)
-> or [Firefox](https://addons.mozilla.org/firefox/addon/sign-in-with-chatgpt/)
-> before connecting. If the hosted chat stays disconnected, confirm the
-> extension is installed and enabled, reload the page, and select **Connect
-> ChatGPT** again.
-
-> The local npm wizard uses its own `localhost:1455` callback instead. If an
-> OAuth extension captures that callback, temporarily disable it during local sign-in,
-> complete the wizard sign-in, and then re-enable it.
-
-While a fresh local login is pending, **Stop sign-in** terminates the helper
-Relmio started before allowing another attempt. Results from older tabs are
-rejected after a replacement starts. If cleanup cannot be confirmed, retry
-remains disabled; close the wizard and OAuth helper, then restart Relmio.
-
-<figure>
-  <img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/examples/hosted-chat-connected.png" alt="Relmio hosted chat showing a connected ChatGPT session and a ready prompt field" width="720">
-  <figcaption>Successful hosted-chat state: the browser extension completed the OAuth handoff and Relmio shows the ChatGPT session as connected.</figcaption>
-</figure>
-
-Use the site's [Install wizard](https://relmio.vercel.app/install) page for a
-clickable macOS/Linux, Homebrew, PowerShell, Command Prompt, and NPX command switcher
-tailored to the current self-hosted n8n and Hostinger VPS setup path.
-
-### Browser interface and theme modes
-
-The hosted site and local wizard keep the original Relmio relay layout while
-adding a compact **System / Light / Dark** appearance control. System follows
-the computer's preference; Light and Dark are remembered on that browser only.
-On desktop, the local wizard keeps progress and sidecar-only safety notes in a
-persistent rail beside the active task; its compact fixed-screen shell avoids
-document scrolling on common laptop screens. On narrow phones, it switches to
-a horizontal progress strip and keeps task scrolling inside the active panel.
-Both the hosted site and local wizard keep Ko-fi support, GitHub stars, and the
-current Relmio version visible beside the appearance control.
-
-## Local Docker endpoints
-
-Choose **Local endpoints** in the browser wizard to install one of these
-Docker services on the same computer as your app:
-
-This local Docker path supports macOS, Linux, and Linux under WSL2. Native
-Windows is not supported in this release because its filesystem permissions do
-not provide the owner-only POSIX mode guarantees used for local credentials.
-The existing VPS/n8n wizard remains available from native Windows.
-
-| Option | Local endpoint | Provider credential | Client type |
-|---|---|---|---|
-| **OpenAI API: compatible clients** | `http://127.0.0.1:12435/v1` by default | Server-side OpenAI Platform API key only | Private local app, SDK, or same-owner development web app |
-| **Codex with ChatGPT: agent clients** | `ws://127.0.0.1:14500` by default | ChatGPT sign-in through Codex | Trusted native Codex/App Server client |
-
-The OpenAI-compatible `/v1` endpoint is powered only by a Platform API key,
-which the wizard seeds over stdin into a private, labeled Docker volume; it
-does not create a host key file. Your app uses a
-separate Relmio capability that the wizard displays once.
-The bearer remains valid until it is rotated. After installation, use
-**Rotate client credential** on the Ready screen to replace only that local
-capability. Relmio shows the new one-time credential before activation, verifies
-the replacement endpoint, and preserves the upstream Platform API key or Codex
-credential/workspace volumes. If replacement cannot be verified, Relmio attempts
-to restore the previous verifier and re-attest service readiness. It does not
-retain the previous raw client credential to replay it during rollback. If
-rollback cannot be confirmed, it targets only the exact managed service for
-shutdown and reports whether that stopped state could be verified. Browser requests
-must come from an exact origin entered during setup; wildcards are not allowed,
-and the capability must never be embedded in a public frontend bundle. Platform
-requests use that API project's billing, credits, limits, and permissions, not
-a ChatGPT subscription.
-
-ChatGPT sign-in powers only the official experimental Codex App Server JSON-RPC
-protocol. It does not expose `/v1`, and Relmio never translates a ChatGPT
-OAuth/session token into a general API credential. OpenAI documents the WebSocket transport as
-experimental and unsupported for production, and it rejects browser-origin
-connections. Use it only with a trusted native client owned by the same person.
-Its capability is high-trust because it can operate the signed-in
-Codex session and files inside the isolated container workspace.
-
-Both services bind exactly to `127.0.0.1`, require the generated capability,
-and mount no host directory or Docker socket. The Codex service gets private
-named credential and workspace volumes. This local path does not connect to a
-VPS or modify n8n.
-
-Read the complete [Local Docker endpoints
-guide](https://github.com/Demonbane18/relmio/blob/main/docs/local-endpoints.md)
-before installing. It includes the wizard steps, client settings, exact-origin
-rules, billing boundary, container isolation, and official OpenAI documentation
-links. The design is documentation-backed engineering guidance, not legal
-advice or an OpenAI approval. Codex for Open Source membership is not treated
-as permission to broaden credential scope or bypass another agreement.
-
-## Quick start
-
-Choose the terminal already on your own computer. Do not run these commands on
-the VPS.
-
-### macOS, Linux, WSL, or Git Bash
-
-```bash
-curl -fsSL https://relmio.vercel.app/install.sh | sh
-```
-
-### Homebrew (macOS or Linux)
-
-```bash
-brew tap Demonbane18/relmio && brew install relmio
-```
-
-### Windows PowerShell
-
-```powershell
-irm https://relmio.vercel.app/install.ps1 | iex
-```
-
-### Windows Command Prompt
-
-```bat
-for /f "delims=" %F in ("%TEMP%\relmio-install-%RANDOM%-%RANDOM%-%RANDOM%.cmd") do @if exist "%~F" (exit /b 80) else curl -fsSL --remove-on-error https://relmio.vercel.app/install.cmd -o "%~F" && set "RELMIO_SELF_DELETE=%~F" && call "%~F"
-```
-
-No Node.js or Git Bash installation is required first. The
-[POSIX](https://github.com/Demonbane18/relmio/blob/main/web/public/install.sh)
-and [Windows PowerShell](https://github.com/Demonbane18/relmio/blob/main/web/public/install.ps1)
-bootstraps, plus the PowerShell-free [Command Prompt](https://github.com/Demonbane18/relmio/blob/main/web/public/install.cmd)
-bootstrap, reuse Node.js 22 or newer when available. Otherwise they show
-staged **Please wait** messages while they download the matching current
-official Node.js 22 runtime to a private temporary directory, verify its
-SHA-256 checksum, run Relmio with npm lifecycle scripts disabled, and remove
-the temporary runtime when the wizard closes. The Command Prompt path runs as
-the current user and does not request administrator access or change Windows
-security policy.
-
-Homebrew is available from the public `Demonbane18/relmio` tap. The WinGet
-command stays hidden until Microsoft accepts its catalog pull request and the
-catalog updates. Until then, use Homebrew or a direct installer above.
-
-Users who already have Node.js 22 or newer can run the npm package directly:
-
-### NPX (requires Node.js 22+)
+## Quick install
 
 ```bash
 npx --yes --ignore-scripts relmio@latest
 ```
 
-Keep the active setup terminal open while using the wizard. If the page says
-the wizard link is incomplete, close that tab and open the full `Local wizard:`
-URL printed by the current terminal instead of refreshing the stripped page.
+The wizard opens a private `127.0.0.1` page, verifies Docker, and requires a
+final confirmation before it writes files or deploys a VPS sidecar.
 
-If the browser does not open automatically, press Enter in an interactive
-terminal to open it again. The printed `Local wizard:` URL remains the fallback
-for a terminal that cannot accept input or when the launcher still cannot open
-a browser.
+## What it can install
 
-### Requirements
+| Option | Contract | Credential boundary |
+| --- | --- | --- |
+| Local OpenAI gateway | OpenAI-compatible `/v1` | Your OpenAI Platform API key |
+| Codex App Server | Experimental JSON-RPC/WebSocket | ChatGPT sign-in and a high-trust local capability |
+| Codex Chat Adapter | Experimental `POST /chat` | ChatGPT sign-in and a bearer for trusted local backends |
+| n8n sidecar | Private Docker-network `/v1` bridge | A local ChatGPT sign-in file, never a host port |
 
-These requirements are for the VPS/n8n path. For a local Docker endpoint, see
-the [local endpoint
-requirements](https://github.com/Demonbane18/relmio/blob/main/docs/local-endpoints.md#requirements).
-
-- On macOS/Linux/WSL/Git Bash: `curl`, `awk`, `tar`, and either `sha256sum` or
-  `shasum`; Git Bash also needs `unzip`
-- On native Windows: Command Prompt uses its built-in `curl`, `certutil`, and
-  `tar` tools; Windows PowerShell 5.1 or PowerShell 7 remains an alternative
-  bootstrap
-- A browser and an eligible ChatGPT/Codex account
-- A self-hosted n8n Docker deployment on a VPS
-- Docker Compose v2, SSH access, and a Docker network shared with n8n
-
-Back up your n8n workflows before granting any tool VPS access.
-
-## What you get
-
-```text
-n8n AI node or HTTP Request
-  -> private Docker endpoint: http://n8n-openai-oauth:10531/v1
-    -> protected ChatGPT/Codex OAuth session
-      -> upstream OpenAI service
-```
-
-Relmio's existing VPS/n8n path provides tested setup instructions for:
-
-- OpenAI Chat Model
-- AI Agent
-- Basic LLM Chain
-- HTTP Request
-
-The local endpoint path described above now supports compatible private apps
-through a Platform-backed gateway and trusted native clients through Codex App
-Server. The broader direction is to add providers and client adapters without
-weakening their authentication boundaries or tying the public product name to
-n8n. SuperGrok/xAI OAuth is a gated feasibility item on the
-[provider roadmap](https://github.com/Demonbane18/relmio/blob/main/docs/roadmap.md);
-it is not currently advertised as supported.
-
-## Visual walkthrough
-
-These current product screenshots use sanitized sample values wherever a setup
-field might otherwise identify an environment. The VPS identity screen shows a
-reserved sample address and sample SSH fingerprint; none of the images includes
-an OAuth token, password, private key, or live wizard session URL.
-
-### 1. Choose a hosted installation method
-
-<img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/setup/00-install-methods.png" alt="Hosted Relmio installation page with macOS/Linux, Homebrew, PowerShell, Command Prompt, and NPX choices" width="720">
-
-The hosted page starts with the installation-method selector. Run the selected
-command locally, not on the VPS.
-
-### 2. Complete the local ChatGPT/Codex sign-in
-
-<img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/setup/01-local-sign-in-ready.png" alt="Sanitized preview-mode sample credential state, not a real saved-login confirmation" width="626">
-
-This image shows a sanitized preview-mode sample credential state; it does not
-confirm that a real local ChatGPT login was saved. During a real fresh login,
-the browser confirmation states that the credentials were saved locally.
-
-### 3. Verify the VPS identity
-
-<img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/setup/02-vps-identity-confirmed.png" alt="VPS identity confirmation with a sanitized sample address and SSH fingerprint" width="720">
-
-Confirm the SSH host fingerprint before authentication. The screenshot uses the
-reserved sample host `vps.example.test`, a sample fingerprint, and a blank
-password field.
-
-### 4. Choose the detected n8n container and network
-
-<img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/setup/03-n8n-detected.png" alt="Detected n8n container and shared Docker network" width="720">
-
-Discovery is read-only; choose the existing n8n container and shared Docker
+ChatGPT sign-in does not become an OpenAI Platform API key. Codex transports
+are not generic `/v1` services and should never be exposed on a LAN or public
 network.
 
-### 5. Review the exact sidecar-only plan
-
-<img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/setup/04-install-plan.png" alt="Exact sidecar-only installation plan, including forbidden actions" width="720">
-
-The plan names the allowed sidecar work and explicitly excludes edits or
-restarts of n8n, a published host port, and a Traefik route.
-
-### 6. Copy the verified n8n settings
-
-<img src="https://cdn.jsdelivr.net/npm/relmio@latest/docs/images/setup/05-bridge-ready.png" alt="Private bridge ready screen with n8n OpenAI credential settings" width="720">
-
-Use these values in an n8n OpenAI credential:
-
-```text
-API Key: local-only
-Base URL: http://n8n-openai-oauth:10531/v1
-Organization ID: leave empty
-Add Custom Header: Off
-```
-
-The `local-only` value is a placeholder required by n8n. It is not an OpenAI
-Platform API key.
-
-For an n8n **HTTP Request** node, use **Generic Credential Type** → **Bearer
-Auth**, name the credential `openai-oauth`, and enter `local-only` as the
-bearer token. Enable **Send Headers** with `Content-Type: application/json`,
-then enable **Send Body** → **JSON** → **Using JSON** and paste:
-
-```json
-{
-  "model": "gpt-5.6-sol",
-  "messages": [
-    {
-      "role": "user",
-      "content": "What is a robot?"
-    }
-  ],
-  "response_format": {
-    "type": "json_schema",
-    "json_schema": {
-      "name": "answer",
-      "schema": {
-        "type": "object",
-        "properties": {
-          "content": { "type": "string" }
-        },
-        "required": ["content"],
-        "additionalProperties": false
-      },
-      "strict": true
-    }
-  }
-}
-```
-
-Use `POST http://n8n-openai-oauth:10531/v1/chat/completions` as the URL. The
-local wizard's **Copy HTTP request recipe** action supplies the same fields.
-Replace the model only if the wizard reports a different ID. The full guide
-has the importable cURL version.
-
-## Important boundaries
-
-- Relmio does not create an OpenAI Platform API key.
-- The OAuth file is a password-equivalent secret; never commit or share it.
-- The sidecar endpoint stays inside the Docker network. Port `10531` must not
-  be published on the VPS host.
-- Relmio never edits, rebuilds, recreates, stops, or restarts the existing n8n
-  container.
-- This is an unofficial community project, not affiliated with or endorsed by
-  OpenAI, xAI, n8n, Hostinger, Telegram, or AppBuildersPH. Provider access,
-  models, limits, and policies can change.
-- Use it only where your account, subscription, provider terms, and applicable
-  policies allow.
-
-## How it works
-
-OpenAI's Codex CLI uses authenticated endpoints at
-`chatgpt.com/backend-api/codex` to run models with a ChatGPT account. Relmio
-uses the same OAuth credential shape through the upstream
-[`openai-oauth`](https://github.com/EvanZhouDev/openai-oauth) helper to expose
-an OpenAI-compatible path without creating an OpenAI Platform API key or
-requiring separate API credits. The upstream service and its access rules may
-change.
-
-### Foundation and attribution
-
-Relmio is built on [`openai-oauth`](https://github.com/EvanZhouDev/openai-oauth)
-by Evan Zhou Dev. That upstream SDK/helper provides the foundation for ChatGPT
-login in local apps and Sign in with ChatGPT flows; Relmio wraps it with n8n
-discovery, verified SSH/SFTP deployment, Docker networking, and
-OpenAI-compatible configuration. Review the upstream project for its own
-license, notices, and supported behavior.
-
-## Known limitations
-
-- Only models supported by Codex are available. The list changes over time and
-  depends on your ChatGPT plan.
-- The CLI `/v1/responses` endpoint is stateless. Callers must send the full
-  conversation history; stateful replay is not provided.
-- Hosted browser sign-in currently supports Chrome and Firefox. Safari is not
-  yet supported by the upstream Sign in with ChatGPT flow.
-
-## Legal
+## ChatGPT sign-in lifetime
 
-Relmio and `openai-oauth` are unofficial, community-maintained projects and
-are not affiliated with, endorsed by, or sponsored by OpenAI.
+ChatGPT/Codex sign-in tokens expire, but the official Codex client refreshes
+them automatically during active use before they expire, so active sessions
+usually continue without another browser login. The official [OpenAI
+authentication documentation](https://learn.chatgpt.com/docs/auth) does not
+publish a fixed 10-day lifetime; do not plan around one. This provider
+credential is separate from Relmio's local capability, which remains valid
+until you rotate it.
 
-ChatGPT OAuth credentials should be treated like passwords. Use your own
-account, keep credentials private, and never pool, share, or redistribute
-access tokens. Do not bypass rate limits, restrictions, or safeguards.
+## Common problems
 
-You are responsible for complying with OpenAI's [Terms of
-Use](https://openai.com/policies/terms-of-use/), [Usage
-Policies](https://openai.com/policies/usage-policies/), and any agreement that
-applies to your account. This project is provided as-is without warranties;
-OpenAI may change or disable the underlying services at any time.
+1. **Docker is not running:** start Docker Desktop or Docker Engine with
+   Compose, then open a fresh wizard. See
+   https://relmio.vercel.app/docs/troubleshooting#docker-is-not-running.
+2. **Authentication fails:** close stale sign-in tabs and use only the newest
+   local wizard URL printed by the active terminal. See
+   https://relmio.vercel.app/docs/troubleshooting#authentication-fails.
+3. **Local image build failed:** check Docker, disk space, and registry
+   connectivity; the browser intentionally does not reveal Docker stderr or
+   local paths.
 
-## Documentation
+Full guides use absolute HTTPS links:
 
-- [Complete GitHub README and manual fallback](https://github.com/Demonbane18/relmio#readme)
-- [Local Docker endpoints](https://github.com/Demonbane18/relmio/blob/main/docs/local-endpoints.md)
-- [Configure n8n AI and HTTP nodes](https://github.com/Demonbane18/relmio/blob/main/docs/n8n-configuration.md)
-- [Troubleshooting](https://github.com/Demonbane18/relmio/blob/main/docs/troubleshooting.md)
-- [Security and limitations](https://github.com/Demonbane18/relmio/blob/main/docs/security.md)
-- [Refresh, upgrade, rollback, and uninstall](https://github.com/Demonbane18/relmio/blob/main/docs/maintenance.md)
-- [Changelog](https://github.com/Demonbane18/relmio/blob/main/CHANGELOG.md)
+- https://relmio.vercel.app/docs/getting-started
+- https://relmio.vercel.app/docs/local-endpoints
+- https://relmio.vercel.app/docs/vps-and-n8n
+- https://relmio.vercel.app/docs/troubleshooting#local-image-build-failed
+- https://relmio.vercel.app/docs/security
+- https://relmio.vercel.app/docs/reference
 
-## License
+Source and issues: https://github.com/Demonbane18/relmio
 
-[Apache License 2.0](https://github.com/Demonbane18/relmio/blob/main/LICENSE).
-See the package
-[NOTICE](https://github.com/Demonbane18/relmio/blob/main/NOTICE) for the
-upstream `openai-oauth` attribution.
+License: Apache-2.0

package/docs/architecture.md

@@ -29,28 +29,37 @@ flowchart LR
   W --> D["Local Docker Engine"]
   D --> G["OpenAI-compatible gateway<br>127.0.0.1:12435/v1"]
   D --> A["Codex App Server<br>127.0.0.1:14500"]
+  D --> H["Codex Chat Adapter<br>127.0.0.1:14501/chat"]
   G -->|"Platform API key"| P["OpenAI Platform API"]
   A -->|"Official Codex sign-in"| C["ChatGPT/Codex service"]
+  H -->|"Official App Server lifecycle"| C
 ```
 
-The two services are intentionally not interchangeable:
+The three services are intentionally not interchangeable:
 
 | Target | Wire protocol | Upstream credential |
 |---|---|---|
 | `openai-api` | OpenAI-compatible HTTP `/v1` | OpenAI Platform API key |
 | `codex-chatgpt` | Official Codex App Server JSON-RPC | ChatGPT sign-in managed by Codex |
+| `codex-chat` | Relmio-specific HTTP `POST /chat` | ChatGPT sign-in managed by Codex |
 
 Relmio never adapts a ChatGPT/Codex credential into the local `/v1` gateway.
 The OpenAI gateway replaces the caller's Relmio capability with the
-protected Platform key only at the upstream boundary. The Codex service keeps
-the native initialization, thread, turn, approval, and event protocol.
-
-Both projects publish exactly one literal `127.0.0.1` binding and require a
+protected Platform key only at the upstream boundary. The native Codex service
+keeps the initialization, thread, turn, approval, and event protocol. The
+adapter invokes that same official lifecycle behind a bounded, read-only
+conversational contract without claiming OpenAI API compatibility. Its model
+sandbox denies network access and uses a root-deny filesystem policy with only
+minimal runtime paths plus `/workspace` readable; `/home/node/.codex` is
+explicitly denied so a model turn cannot read the persisted ChatGPT session.
+
+Each project publishes exactly one literal `127.0.0.1` binding and requires a
 generated bearer capability. Their managed roots are
-`~/.relmio/local/openai-api` and `~/.relmio/local/codex-chatgpt`. The Codex
-credential and workspace use private named Docker volumes; no host directory
-or Docker socket is mounted. See [Local Docker endpoints](local-endpoints.md)
-for setup and trust limitations.
+`~/.relmio/local/openai-api`, `~/.relmio/local/codex-chatgpt`, and
+`~/.relmio/local/codex-chat`. The Codex credentials and workspaces use
+target-specific private named Docker volumes; no host directory or Docker
+socket is mounted. See [Local Docker endpoints](local-endpoints.md) for setup
+and trust limitations.
 
 Before installation, Relmio resolves the selected Docker context to a local
 Unix socket and pins that exact socket on every later Docker command. Remote
@@ -182,3 +191,6 @@ sidecar authenticates upstream with the mounted OAuth file.
   or mismatched managed identity blocks local mutation.
 - A Codex login failure returns a sanitized status without returning App
   Server output or ChatGPT tokens.
+- A chat adapter request with a browser Origin, invalid bearer, malformed body,
+  protocol overflow, timeout, or failed turn is rejected with a sanitized
+  response and its App Server helper is terminated.

package/docs/faq.md

@@ -0,0 +1,45 @@
+# Frequently asked questions
+
+## Does a ChatGPT plan include a Platform API key?
+
+No. The OpenAI-compatible `/v1` local endpoint uses a user-supplied OpenAI
+Platform API key, with Platform billing. ChatGPT sign-in is only for the
+experimental Codex paths and is never represented as generic `/v1`
+authorization.
+
+## How long does a ChatGPT/Codex sign-in token last?
+
+ChatGPT/Codex sign-in tokens expire, but the official Codex client refreshes
+them automatically during active use before they expire, so active sessions
+usually continue without another browser login. The official [OpenAI
+authentication documentation](https://learn.chatgpt.com/docs/auth) does not
+publish a fixed 10-day lifetime; do not plan around one. This provider
+credential is separate from Relmio's local capability, which remains valid
+until you rotate it.
+
+## Can I expose the local endpoints to my network?
+
+No. Relmio's local endpoint plans bind only to `127.0.0.1`. Do not reverse
+proxy or port-forward the Codex targets. They are intended only for the local
+machine and their stated trusted-client boundary.
+
+## Why does the Chat Adapter reject browser requests?
+
+The adapter deliberately has no CORS support and rejects requests with an
+`Origin` header. It is for a trusted local backend or development server, not
+browser bundles. The in-wizard tester is the one narrow exception: the browser
+talks only to the setup-token-protected wizard, which makes the server-side
+adapter request without an `Origin` header.
+
+## Is the in-wizard tester end-to-end encrypted?
+
+No. It uses an expiring in-memory RSA-OAEP key to reduce accidental credential
+transit and storage exposure between the wizard page and local server. It does
+not protect against a compromised browser, extension, or local machine, and it
+does not provide encryption at rest or end-to-end encryption.
+
+## Can Relmio modify my n8n deployment?
+
+No. It deploys a distinct sidecar only after you approve the displayed plan.
+It does not edit existing n8n Compose files or images and never publishes port
+`10531` on the VPS host.