Venom Tasks settings
Settings in Venom Tasks has a left nav with your own Preferences and Notifications, workspace pages for teams, members, labels and import, and one page per team with a danger zone.
Open Settings from the sidebar (/tasks/settings). A nav on the left holds the pages; on a phone it sits above
the page.
| Group | Pages |
|---|---|
| Account | Preferences, Notifications |
| Workspace | General (/tasks/settings: the plan and the teams), Members and agents (/tasks/settings/members), Labels (/tasks/settings/labels), Import (/tasks/settings/import). Agents opens the Agents section of General; Templates and Recurring issues open those sections on the first team's page (each team page has its own; see Templates) |
| Teams | One page per team |
Preferences
/tasks/settings/preferences. These are yours alone and follow you to any browser. Each control saves as you change it.
| Setting | What it does |
|---|---|
| Default home | Where /tasks opens: My issues (the default), the view you had open last, the Inbox, or your first team's issues |
| Display density | Compact fits more on the screen in lists, boards and the Inbox |
| First day of the week | Monday, Sunday or Saturday. "Next week" in Snooze starts on it |
| Opening an issue | Enter on a focused issue opens its page, or peeks beside the list |
| Copying a branch name starts the issue | On by default. Mod+Shift+. also assigns the issue to you when nobody has it and moves it to its team's first started state; Mod+Z undoes both (GitHub) |
| Send comments with | ⌘/Ctrl+Enter (the default) or Enter. With Enter, Shift+Enter starts a new line, and inside a list or a code block Enter keeps editing (send with ⌘/Ctrl+Enter there). On a phone the return key always makes a new line and the button sends |
| Assign new issues to me | Off by default. The new-issue form starts with you as the assignee (you can still change it) |
| Assign issues I start to me | Off by default. Moving an unassigned issue to a started status assigns it to you. The server does it, so the board, the API and your MCP client behave the same; an agent starting an issue never triggers it. The issue's activity says "took it on starting it (personal setting)" |
| Turn typed emoticons into emoji | On by default. :) and a space becomes 🙂 in descriptions, comments and notes; Backspace right after puts the text back (the table) |
| Local data | Clear local data deletes the copy of your lists and recently opened issues that this browser keeps so Tasks opens at once (Fast reopening and offline reading). It applies to this browser only; Tasks saves a new copy the next time you open it |
Over the API these are commentSend (mod-enter or enter), auto: { assignOnCreate, assignOnStart } and
emoticons (true or false) on PUT /tasks/api/preferences (a partial object).
Notifications
/tasks/settings/notifications. A table: each row is a kind of event, each column a place it reaches you.
| Row | Events |
|---|---|
| Assigned to you | assigned, unassigned |
| Mentions | mentioned in an issue or a comment |
| Comments | new comments on issues you follow |
| Status and priority changes | status, urgent priority, a related issue linked |
| Agent updates | an agent's comments and changes, and an agent's expired claim. An agent that assigns or mentions you still reaches you with this off |
| Due dates and SLAs | due soon, SLA at risk, SLA breached |
| Reminders | a reminder comes due, a snoozed issue wakes up |
| Project updates | an update you lead or own is due, and the two nudges after it (update schedule) |
| Column | Default | What it does |
|---|---|---|
| Inbox | every row on | What reaches your Inbox and the bell. Turning a row off stops new ones; the ones already there stay |
| Push | on for Assigned, Mentions, Due dates and SLAs, Reminders, Project updates | A notification on your phone or computer at once. A tap opens the issue (in the installed app on a phone) |
| on for Assigned and Mentions | One email, only for what you have not read in Venom by then |
Push and Email never carry what your Inbox row filters out, and your own actions never notify you. The switch above the Push and Email columns turns that channel off without losing the rows. Pause push and email stops both at once; your Inbox keeps working.
Push. Turn on Push on this device in each browser or phone you want notified (your browser asks first; on an iPhone or iPad, add Venom to the Home Screen first, iOS 16.4 or later). Send a test checks it arrives. One browser serves both Tasks and Agents. A push shows the issue's ID and title ("ENG-12 · Fix login") and what happened ("sam assigned it to you"); with Hide issue titles on the lock screen it shows the ID alone. It never holds a comment's text or the description. Several pushes about one issue replace each other instead of stacking. The push column is hidden when the install has no VAPID keys (see For operators).
Email. A notification waits unread for your delay (5, 10 (the default) or 30 minutes, 1 hour, or Twice a day
at 9:00 and 16:00 in your time zone), then goes in one email with everything else still unread, grouped by issue,
newest first, at most 30 issues. Read it in Venom first and no email goes. Email is written in your app's language.
It goes out through the mail provider your organisation connected in Support settings; without one there is no
email, nothing waits for one, and the Email column says so. Every email has a link per kind of event in it that turns
that one row's email off in one click, with no sign-in, and a Manage notifications link here; mail clients' own
unsubscribe button (List-Unsubscribe, RFC 8058) turns off that row, or all Tasks email when the message mixed
several. No images, tracking pixels or rewritten links.
Safeguards. Each notification is pushed at most once and emailed at most once, even with several sweeps at the same time. Whether you may still open the issue is checked when it is sent: leave a private team and its issues stop reaching you. A member gets at most 30 pushes and 6 emails an hour (40 emails a day), an organisation at most 3,000 pushes and 1,000 emails an hour; past that a push is dropped (the Inbox still has it) and an email waits. A failed email or push is tried again with growing waits, at most five times.
Quiet hours. Off by default. On, push and email wait from the start time to the end time in your time zone (22:00 to 08:00 unless you change them), then one push sums up what came in; your own reminders still push. See Quiet hours, and Mute a thread for one noisy comment thread.
For the whole organisation. Owners and admins see two more switches: Allow push notifications and Allow notification email. Off means off for every member, whatever their own settings.
The Email digest switch emails your weekly Pulse digest on Monday, as well as showing it in the bell. It goes out through the email provider your organisation connected in Support settings; without one, nothing is sent and the digest on the Pulse page says why. Off by default.
Agents read and change your notification settings with tasks_notification_settings_get and
tasks_notification_settings_update (only yours, and only with write access); an admin's agent turns a channel off
for the organisation with tasks_notification_channels_update.
Workspace pages
/tasks/settings/members, Members and agents: everyone in the organisation (anyone can be an issue's assignee) and the agents that take delegated work: MCP clients, Venom daemons and sessions. People are invited and removed in Venom Cloud's members settings (/settings/members). An admin creates a service agent here (New agent: a name and a team key) and can rename one; Edit access narrows what it may do. An admin can also give a service agent a computer (computers_agent_set,kind: "tasks"): an issue delegated to it then runs on its own browser on one of your daemons, and its questions and result are comments by the agent. See Agents on computers./tasks/settings/labels, Labels: the workspace labels and label groups every team can use. An issue takes one label per group. A team's own labels are on its team page./tasks/settings/updates, Project updates: the workspace's update cadence (owners and admins; weekly on Friday at 14:00 until changed) and, per team, whether its projects expect updates (off by default). See Update schedule.
Team pages
/tasks/settings/teams/{key} has the team's name and key (its issue prefix), Parent and subteams (move the team
under another one, its subteams, a new subteam), Workflow (states and Triage), Labels, Estimates, Cycles, its Members
(read only), and a Danger zone. A subteam's Estimates and Cycles sections have a switch to follow the parent team's
scale and schedule (Subteams). The same settings, and a quick settings dialog,
are a click away from the sidebar's team menu (The team menu).
Teams are managed for the whole organisation in Settings → Teams: the team's name, subteams, members and roles, the private flag, leaving the team and retiring it (archiving: read-only, history kept, out of the sidebar). Each section that moved there links to it. The key (the issue prefix) is the org team's: change it here or there, and old identifiers keep working. What stays in Tasks is what only Tasks has: the workflow and the rest of the settings above, New team under General (a Tasks team with its own workflow, optionally a private subteam), and:
- Workflow automations (in Workflow; Workflow and triage automations below).
- Delete team removes the team and all its issues for good, after you type the team's key. A team with issues has to be archived first, so deleting is always a second step.
Changing a team's workflow and settings needs the team's admin role (or an admin role on a team above it) or an org role that manages Tasks (owner or admin); the page is read only for everyone else. A private team's page is "not found" to people who can't see the team.
Workflow and triage automations
Each is off until the team turns it on, in the team's Workflow section. Whatever an automation changes writes a line in the issue's activity saying which rule did it ("closed it with its parent (team automation)"), and can be changed back like any other edit. None of them runs on history brought in by an import (the Linear import replays status changes).
| Setting | What it does |
|---|---|
| Close open sub-issues when the parent is done | When a parent issue is completed, its open sub-issues move to their team's first completed status, as Venom Tasks. Each of those closes its own open sub-issues when its team has the switch on too. Already closed or archived sub-issues are left alone. A canceled parent changes nothing. With Close parent issues also on, closing the last sub-issue closes the parent, which doesn't close anything again |
| When an issue moves to a later status | Keep its place (the default), Put it first or Put it last in its new status. "Later" follows the workflow's order (type, then position). With first or last on, moving back to an earlier status puts the issue first. It sets the issue's manual order, so it shows in lists ordered manually |
| Who handles new triage issues | Shown when Triage is on. Nobody (the default), Notify (everyone on the rota gets a "New in Triage" item in their Inbox, under the Assigned notifications) or Assign in turn (each issue that lands in Triage goes to the next person on the rota, round robin). The rota lists people of the team (or of a team above it); someone who left the team or the organisation, or can no longer see a private team, is skipped. An issue that already has an assignee keeps it and doesn't use a turn. With Assign in turn, the page and the team's Triage list show who's next ("On triage: Sam") |
| Ask for a priority before accepting | Accepting (1) an issue without a priority asks for one first; picking one accepts. The API and MCP refuse it with 400 priority_required unless the request sends priority |
Triage duty runs once per arrival: when an issue is created in Triage, or moved into it by a person, an agent or the API, after the team's triage rules. An issue a rule sends on to another team's Triage meets that team's duty, once; a rule's own changes never start it. Two issues landing at the same moment take two turns. Duty is per team: a subteam doesn't inherit its parent's rota, so set one on each team that triages.
Over the API and MCP (tasks_team_update), these are autoCloseChildren, orderOnState (first, last, keep),
triageDuty (none, notify, assign), triageDutyUids (member uids in rota order, at most 100) and
triageRequirePriority; a team also reads triageOnDuty, who the next issue goes to.
API
| Endpoint | Does |
|---|---|
GET /tasks/api/preferences |
Your settings, with defaults filled in |
PUT /tasks/api/preferences |
A partial update: {density: "compact", inbox: {comments: false}}. Unknown keys and values are refused with 400 |
GET /tasks/api/notifications/settings |
Your inbox and delivery switches, and channels: whether this install and org can push and email (available, org, reason: no_vapid, no_mail_provider, org_off) |
PUT /tasks/api/notifications/settings |
A partial update of inbox and delivery (paused, hideTitles, push: {on, events}, email: {on, delay, events}, locale, timeZone, quiet: {on, from, to} with times as HH:MM). delay is 5m, 10m, 30m, 1h or digest |
PUT /tasks/api/notifications/channels |
The organisation's switches {push, email}. Owners and admins; audited |
GET, POST, DELETE /tasks/api/notifications/push |
This browser's push subscription (the same ones as Agents'); POST …/push/test sends a test |
GET, POST /tasks/api/notifications/unsubscribe?t=… |
The email's one-click unsubscribe: GET shows a page and changes nothing, POST turns that one switch off. The signed link is the only credential |
PATCH /tasks/api/teams/{id} with {archived: true} |
Retire; false restores |
DELETE /tasks/api/teams/{id} |
Delete a retired or empty team |