Data Collection
We understand that when you use a bot, there’s a certain level of privacy you expect.
That is why we provide this information below, to clear up any concerns.
Data We Collect
session ids
What: | An MD5 hash of your user ID and the channel ID |
When: | Glyph is explicitly mentioned via DM or @mention |
Where: | DialogFlow requests and message logs |
Why: | Used with DialogFlow to ensure any follow up intents are personalized to you |
messages
What: | A copy of your message contents, your session ID and what skill Glyph interpreted it to be |
When: | Glyph is explicitly mentioned via DM or @mention, only |
Where: | Temporary log on Heroku, logs generally disappear within 24 hours |
Why: | To make sure Glyph is responding appropriately or see what messages caused an error |
configurations
What: | The server ID and the configuration settings |
When: | A custom configuration is set by a server admin (deleted when Glyph is removed from the server) |
Where: | A PostgreSQl database managed by Heroku |
Why: | To access and store a server’s unique configuration |
server info
What: | The server name, ID, member count (humans and bots) |
When: | Glyph is added or removed from a server |
Where: | A private Discord channel (sent via webhooks) |
Why: | To see what servers Glyph has been added and removed from and see which servers are considered bot farms |
username history
What: | The old and new usernames of users |
When: | A user changes their username |
Where: | The auditing channel of any server that has Glyph configured to track username changes |
Why: | So server admins can remember who people are even if they change their name |
Data We Don’t Collect
Obviously anything we don’t collect is not listed above but here’s a list of common things we don’t collect.
- Deleted messages
- Message edits
- The username, ID, or discriminator of who sent feedback
- Audit logs (once they are sent, they are forgotten)
- Ban reasons
- Kick reasons