collabcentral/README.md
Joseph B. McQueen 44568044e8 Phase R1: split per-user data into gitignored authorized.json
Every add/remove of a favorite group or authorized user was rewriting
config.json — the same file that carries structural bot metadata and
was committed to git. This split ends the git-noise and lets ops
deploy fresh installs without a pre-populated user list.

Split
- config.json (committed) stays structural: server, per-bot labels,
  integration + service-account ids, languages.
- config/authorized.json (gitignored) is the new mutable source of
  truth: { admins: [personId...], bot: { <appName>: { <personId>:
  { id, displayName, email, avatar, groups: [...] } } } }.
- Seeded authorized.json with the current admins list and all
  authorized users (3 on novi, 4 on techupdates) so this commit is
  a pure move — no data lost, no downtime.

Helpers (lib/helpers.js)
- New getAuthorizedEntry(authorized, app, id) as the single lookup
  point every consumer goes through, so nullability is uniform.
- isAuthorized() gains an authorized-doc arg (pure signature stays
  testable): fails closed when the doc is missing / partially
  loaded, so a broken deploy grants no access.
- isAdmin() now reads authorized.admins instead of config.admins.

Runtime (index.js)
- loadAuthorized() with an ENOENT fallback to { admins: [], bot: {} }
  so a fresh deploy can bootstrap via the admin page instead of
  requiring a hand-crafted authorized.json.
- All 8 previous config.webex.bot[app].authorized sites (favorites
  read/add/remove, admin list/add/delete, isAuthorized) now go
  through the authorized doc.
- Every mutation writes to config/authorized.json instead of
  config/config.json.

Latent-bug fixup (uncovered while smoke-testing this refactor)
- The /user/:scope/:action fallthroughs used res.status(4xx)
  without .send(...), so unknown scopes / unauthorized callers got
  a hung request instead of a response. Added ".send(...)" bodies
  so the response actually completes.

Docs + tests
- README updated: new "Authorized users" step in "Adding a new bot",
  updated file-layout section, docker mount list adds
  authorized.json.
- Test suite expanded from 48 → 53 with a new getAuthorizedEntry
  group and the existing isAuthorized/isAdmin cases reshaped for
  the new signatures.

Smoke tested the auth matrix end-to-end (admin + non-admin + signed-
out across /info, /admin/users, /user/groups/list): every path
returns the expected code and body.

Co-authored-by: Cursor <cursoragent@cursor.com>
2026-07-01 21:06:44 -04:00

182 lines
6.8 KiB
Markdown

# CollabCentral
A multi-bot mass-messaging service for Webex. Each bot has its own token,
avatar, label, and per-user access list, and users can only see the bots
and jobs they're authorized on.
Under one server you can host any number of bots (Novi, TechUpdates, …) —
each one exposes the same UI at:
- `/CollabCentral/<botName>/sendMessage.html` — compose and send.
- `/CollabCentral/<botName>/monitorJobs.html` — track running / scheduled / completed jobs.
- `/CollabCentral/<botName>/jobDetail.html?jobId=…` — detailed per-job view.
## Requirements
- Node.js 20+.
- A single Webex integration used for OAuth on all bots.
- A Webex service account (used server-side for group/people lookups; its
refresh token is stored in `config/token.json` and rotated automatically).
- One Webex bot per app you want to host.
## Quick start (local)
1. Copy `.env.example` to `.env` and fill in the values:
```
SERVER_PORT=1450
CRON_TIMEZONE=America/New_York
OAUTH_CALLBACK_URL_TEMPLATE=https://bot.example.com/CollabCentral/:app/oauth
WEBEX_INTEGRATION_CLIENT_ID=...
WEBEX_INTEGRATION_CLIENT_SECRET=...
WEBEX_SERVICE_ACCOUNT_CLIENT_ID=...
WEBEX_SERVICE_ACCOUNT_CLIENT_SECRET=...
GOOGLE_TRANSLATE_API_KEY=...
```
2. Copy `config/botTokens.example.json` to `config/botTokens.json` and put
in the bot tokens for each app:
```json
{
"novi": { "token": "…", "enabled": true },
"techupdates": { "token": "…", "enabled": true }
}
```
3. Create `config/token.json` with the initial service-account OAuth tokens
(access + refresh). After the first refresh, the cron job rewrites this
file automatically.
4. Install dependencies and start:
```
npm install
npm start
```
`npm start` uses `node --env-file=.env`, so no external process manager is
required to load env vars during local development.
## Adding a new bot
Say you want to add a bot called `alerts`.
1. **Webex bot** — create the bot at developer.webex.com. Copy its access token.
2. **Bot token** — add an entry to `config/botTokens.json`:
```json
"alerts": { "token": "<bot access token>", "enabled": true }
```
Setting `enabled: false` will keep the bot listed but return 404 for all
`/CollabCentral/alerts/*` routes.
3. **Bot metadata** — edit `config/config.json` and add a structural entry
under `webex.bot`:
```json
"alerts": { "label": "Ops Alerts" }
```
4. **Authorized users** — add per-user data to `config/authorized.json`
(gitignored; created automatically at first admin action if it doesn't
exist):
```json
{
"admins": ["<webexPersonId>"],
"bot": {
"alerts": {
"<webexPersonId>": {
"id": "<webexPersonId>",
"displayName": "…",
"email": "…",
"avatar": "…",
"groups": [ { "name": "…", "alias": "…", "id": "<webexGroupId>" } ]
}
}
}
}
```
Only person IDs listed under `bot.<appName>` can sign into that bot's UI.
Anyone listed under `admins` (any personId) can manage authorized users
across every bot via the in-app Admin page — much easier than hand-
editing this file, but the file is the source of truth.
4. **Icons** — drop `alerts.png` and `alerts.ico` into `html/` (they're
served by the per-bot static mount).
5. **OAuth redirect URI** — in the Webex integration used for OAuth, register
`https://<your-host>/CollabCentral/alerts/oauth` as an allowed redirect
URI. Without this step, users get an OAuth error on login.
6. Restart the server (or the container). The bot's avatar is fetched once
at startup via Webex `/people/me` using the token you supplied in step 2.
## Docker
`Dockerfile` builds a stateless image. Config, tokens, uploads, and jobs
data are expected to be mounted at runtime.
```
docker build -t collabcentral .
docker run -d \
--name collabcentral \
--env-file /path/to/.env \
-v /path/to/config:/usr/src/app/config \
-v /path/to/uploads:/usr/src/app/uploads \
-p 1450:1450 \
collabcentral
```
The `config/` mount should contain at minimum `config.json`, `botTokens.json`,
`token.json`, `languages.json`, `authorized.json`, and (if you want to
preserve state) `jobs.json` and `userPrefs.json`.
## File layout
Committed to the repo:
- `index.js` — Express server, OAuth, cron jobs, Webex API integration,
and the send queue.
- `html/` — frontend (sendMessage / monitorJobs / jobDetail pages, per-bot
icons, shared JS libraries).
- `config/config.json` — structural: server settings, per-bot labels,
integration ids. No per-user data lives here anymore.
- `config/languages.json` — supported translation languages.
- `config/botTokens.example.json` — template for `botTokens.json`.
- `.env.example` — template for `.env`.
- `Dockerfile`, `.dockerignore`, `.gitignore`, `package.json`,
`package-lock.json`.
Runtime state (gitignored, not committed):
- `.env` — secrets and deployment-specific URLs.
- `config/botTokens.json` — per-bot access tokens.
- `config/token.json` — service-account OAuth tokens (rotated by the
refresh cron).
- `config/authorized.json` — admins list + per-bot authorized users
(with their favorite groups). Edited at runtime by the admin page
and the favorites picker on the compose page.
- `config/jobs.json` — building / running / scheduled / completed jobs.
- `config/userPrefs.json` — per-user preferences (language, etc.).
- `uploads/` — CSVs of recipient IDs and any attached images.
## Testing
Unit tests for the pure helpers extracted into `lib/helpers.js` run under
Node's built-in test runner — no additional dev dependencies required.
```
npm test
```
The suite covers bot enable/disable gating (`getBotToken`, `isBotEnabled`,
`getBotConfig`), per-bot draft isolation (`buildingKey`, `jobsForApp`), the
authorization check (`isAuthorized`), OAuth URL construction (`buildAuthUrl`,
`getOAuthRedirectUri`), the retention filter (`cleanCompletedJobs`), and the
duration formatter (`msToTime`). Adding a new helper? Add it to `lib/helpers.js`
and cover it in `test/helpers.test.js` — keeping the state-carrying wrappers
in `index.js` thin means each helper can be tested without booting the server.
## Behavior notes
- **Completed-job retention**: the daily cleanup cron (default `01:10` in
`CRON_TIMEZONE`) drops entries from `jobs.completed` older than 30 days.
Change `COMPLETED_RETENTION_DAYS` in `index.js` to adjust.
- **Send concurrency**: a shared `p-queue` limits outbound Webex sends to
10 in-flight requests.
- **Rate limiting**: `fetchWithRateLimit` transparently retries on Webex 429
responses honoring `Retry-After`.
- **Per-bot job isolation**: draft jobs are keyed by `cookieId + appName`,
and job list / detail endpoints filter by `appName`, so authorized users
of one bot never see another bot's jobs.