Your projects.
Your data. Your Donna.

Add text that Donna should know

1. Sign in and open Projects & data. Select a project from Active project, or choose New project to start with empty storage.
2. Under All data → Knowledge, click Add knowledge.
3. Enter a title, category, and your text. You can also load a text file (.txt, .md, .csv, or .json, under 200 KB), then review its contents.
4. Click Save knowledge. The entry is saved in the database and becomes available to bots assigned to this project. Refresh the page to see that it remains saved.

Choose which bot uses a project

1. Select your project and open Connections.
2. For a bot already in your workspace, click Use an existing bot, choose it from the list, and click Use selected bot. No bot is chosen automatically.
3. For a new bot, open the verified @BotFather in Telegram, send /newbot, and follow the prompts. Copy the bot token.
4. Back in the project, click Connect bot, enter its label and token, and connect it. Add the bot to your Telegram group and follow the group setup prompts.

Screenshot: where to paste a Telegram bot token

Replace an expired or revoked token

In BotFather, open /mybots, select the same bot, and get a fresh API token. In Telecore, open Telegram bots → the bot’s actions → Replace token. Paste the new token and click Validate & replace. The bot keeps its project and records. A paused bot stays paused until you resume it.

Changing a bot’s project changes the knowledge and connections it uses. Records already saved in its previous project stay there. Allow a short runtime refresh for the bot to pick up the change.

Token instructions: Telegram’s BotFather guide.

Browse records and graphs

All data lets you search knowledge, tasks, documents, meetings, conversation history, people, reminders, and imported Trello cards. Click a record to inspect it. Choose Analytics for task activity, completion status, and collection counts.

Use Export data to download the selected project’s records as JSON. Connection credentials are excluded.

Bring a board into your project

1. Open Trello’s app admin, create or select your Power-Up, then open Trello Auth → Generate a new API Key. Copy the key.
2. Request a read token by replacing YOUR_API_KEY in the address below. Open the completed address in your browser, approve access, and copy the returned token. This example expires in 30 days.

https://trello.com/1/authorize?expiration=30days&scope=read&response_type=token&key=YOUR_API_KEY

3. In Telecore, choose Projects & data → your project → Connections → Connect Trello.
4. Paste the API key, API token, and Board URL or ID. Click Connect & sync.
5. Open All data → Trello to inspect imported cards. Click Sync now after changes in Trello. Replace the token when it expires.

Each connection belongs to one project. Sync reads your board; it does not edit cards in Trello. Donna can use the imported snapshot when answering in that project.

Credential and token steps: Trello’s authorization guide.

Which key goes where?

Enable AI responses

The platform administrator creates an OpenRouter key, pastes it into Settings → Providers, and clicks Save OpenRouter key. The provider account needs available credits. Workspace owners can then select the available AI model in their settings.

For deployment configuration, the app also supports OPENROUTER_API_KEY in Railway’s service variables. A key saved in Telecore’s provider settings takes precedence.

Provider reference: OpenRouter authentication.

Keep text saved on Railway

Telecore stores project text, accounts, tasks, and connection settings in SQLite. The production database lives at /data/tasks.db on the service’s attached Railway volume. All projects use this durable database with separate project storage and workspace access checks.

1. In Railway, open the existing Telecore project and select the app service.
2. Keep its existing volume attached at /data. For a fresh installation, add a volume to the app service and set this mount path.
3. Open the service’s Variables tab. Choose New Variable, enter the name and value, and apply the changes with a deployment.

Screenshot: where to enter Railway variables

Screenshot: setting a volume mount path

Deployment checks

The start command is python start_service.py. It checks the volume path, takes a database backup, runs migrations, and starts the dashboard and Telegram worker. Railway checks /api/health. Run one service replica with this SQLite volume and polling worker.

Startup backups are stored under /data/backups; the five most recent startup backups are retained. They share the volume. Use Railway’s volume backup controls for a separate recovery point. Keep the original volume attached when redeploying.

Hosting reference: Railway volumes · Railway variables.

Get a Railway API token

A Railway token lets deployment tools manage Railway. You do not need one to add knowledge or connect bots in Telecore.

1. For automation limited to this project, open Railway → Project settings → Tokens. Create a token for the production environment.
2. Copy it into your CI or deployment tool’s secret store as RAILWAY_TOKEN. The Railway CLI reads that variable.
3. For account or workspace access, open Railway account tokens, create a token with the appropriate workspace scope, and use RAILWAY_API_TOKEN for CLI automation.

Screenshot: Railway’s account/workspace token form

The running Telecore app needs its database and provider credentials. It has no Railway token field because deployment access belongs in your deployment tools.

If something is missing