Export from Venom Tasks
Take your data out of Venom Tasks as CSV for any list, project or initiative, or as a ZIP of CSV and JSON for the whole workspace, with the exact columns and the JSON schema, who may export what, and how an agent does it.
A team should be able to leave a tracker as easily as it joined. Venom Tasks exports in two ways: any list as a CSV in one click, and the whole workspace as a ZIP that another tool, or Venom's own importer, can read.
Export a list
Open the ⋯ menu next to Insights on a team's list, on My issues or on a saved view and choose Export as CSV (or press ⌘K and run Export issues as CSV). The file holds exactly the issues the list shows: its filters, the text search, Match any, the insight filter from a chart click, the include-subteams switch, whether sub-issues are shown, and its ordering. Columns you hide on screen are still in the file; the columns are fixed so a spreadsheet or a script can rely on them. The rows come in exactly the order the list shows them, reversed when the list is (Display > Reverse ordering; the manual order is never reversed) and with closed issues last, most recently closed first when that switch is on. Grouping is not applied: like Linear's, the CSV is one flat list, sorted by the ordering and its direction, not group by group.
The same file comes from the API: GET /tasks/api/issues?format=csv takes the list's query parameters (team,
assignee, label, project, cycle, stateType, priority, sla, q, view, filter, orderBy, reverse (1: the ordering the other way round), closedLast (1: closed issues last), subteams,
includeArchived, includeSubIssues) and streams the file: the server never holds more than a page of rows, so an
export of 50,000 issues needs no more memory than one of fifty.
| Who | Most issues in one export |
|---|---|
| A member (any role that can open Tasks) | 5,000 |
An owner or admin (or a key with tasks:admin) |
50,000 |
Past the limit the export is refused with a message that says to narrow the list or export the workspace.
The issue columns
The header row is English in every language, in this order. Dates are ISO 8601 in UTC.
| Column | What it holds |
|---|---|
ID |
The identifier, ENG-12 |
Team |
The team's key |
Title, Description |
The title; the description as Markdown |
Status, Status type |
The status name; its type (triage, backlog, unstarted, started, completed, canceled) |
Priority |
Urgent, High, Medium, Low or No priority |
Estimate |
Points, empty when not estimated |
Assignee, Assignee email |
The owner's name and address |
Agent |
The delegate agent's name |
Creator |
Who created it |
Labels |
; separated; a label in a group is Group › label |
Project, Milestone |
Their names |
Cycle |
The number, then the name when it has one: 4 Sprint 4 |
Parent ID, Sub-issue count |
The parent's identifier; the number of open sub-issues |
Due date |
YYYY-MM-DD |
Created, Updated, Started, Completed, Canceled, Archived |
Timestamps, empty when not set |
SLA status |
onTrack, atRisk, breached, met, or empty |
Customer requests |
How many customer requests you may see on it |
Linked PRs |
Pull request addresses, ; separated |
URL |
The issue's address |
A cell is written as RFC 4180 CSV (quoted when it holds a comma, a quote or a line break), the file starts with a byte
order mark so Excel reads it as UTF-8, and a cell that starts with =, +, -, @, a tab or a carriage return gets
a leading apostrophe, so a title can never run as a formula in a spreadsheet. Venom's importer takes that apostrophe off
again.
Projects, initiatives and members
The ⋯ menu on Projects and on Initiatives exports that table (GET /tasks/api/projects/export,
/initiatives/export). Both include archived ones and only what you may see.
- Projects:
Name,Status,Health,Priority,Lead,Members,Teams,Initiatives,Start,Target,Progress %,Issues (open/total),Latest update(date, health, the first 500 characters),Created,Completed,URL. - Initiatives:
Name,Status,Owner,Target,Projects,Latest update,URL. - Members (owners and admins only, Settings → Export,
GET /tasks/api/members/export):Name,Email,Role,Teams,Joined.
Export the workspace
Settings → Export → Export everything (/tasks/settings/export, owners and admins) starts a background export. It writes one ZIP:
| File | What it holds |
|---|---|
issues.csv |
Every issue, archived included, in the columns above |
projects.csv, initiatives.csv |
As above |
cycles.csv |
Team, Number, Name, Starts, Ends, Completed, Issues, Completed issues |
labels.csv |
Name, Group, Team, Color, Description, Issues, Archived |
members.csv |
As above |
venom-tasks.json |
Everything, below |
It runs in steps on the server, so you can close the page; the export continues, the page shows progress while it is open, and when it finishes it appears in the bell, Your workspace export is ready. The download is for you alone and stops working 24 hours later, when the file is deleted. One export runs at a time per org, an org can start three a day, and the page lists your past exports with Download and Delete. Attachment files are not included, only their names; download them from the issues (or from Notes, where imported files live).
venom-tasks.json
One JSON object, "format": "venom.tasks-export", "version": 1. A reader should check both, ignore keys it does not
know and treat a larger version as a different format. Ids are Venom's and stay stable; identifiers (ENG-12) are
what people use. Every list is ordered the way the app shows it, and the issues come last in the file so a reader can
take the small tables first.
| Key | Holds |
|---|---|
format, version, exportedAt, exportedBy, note |
The header; note says what is left out |
teams |
The team objects of GET /tasks/api/teams (key, name, parent, privacy, settings, estimate scale, cycles settings) |
states |
An object { <teamId>: [ { id, name, type, color, position } ] } |
labels |
Labels and label groups (isGroup, parentId, teamId, color), archived included |
cycles |
Every team's cycles (number, name, startsOn, endsOn, completedAt, progress) |
projects |
Projects with status, lead, members, team ids, dates, progress and health |
milestones |
{ id, projectId, name, description, targetDate } |
projectUpdates, initiativeUpdates |
{ id, projectId | initiativeId, health, body, authorKind, authorId, createdAt } |
initiatives |
Initiatives with status, owner, target and projectIds |
members |
{ id, name, email, role, teams[], joinedAt } |
issues |
One object per issue, below |
An issue object: id, identifier, teamId, number, title, description (Markdown), state { id, name, type },
priority (0 to 4), estimate, assignee { id, email }, delegateAgentId, creator { kind, id, name },
labels [ { id, name, group } ], parentId and parentIdentifier (sub-issues are the issues that name a parent),
projectId, milestoneId, cycleId, dueDate, the timestamps (createdAt, updatedAt, startedAt, completedAt,
canceledAt, archivedAt), slaStatus, and the lists:
comments:{ id, parentId, author { kind, id }, body, createdAt, editedAt, resolvedAt }relations:{ type, issueId, identifier }withtypeone ofblocks,blocked_by,related,duplicate_of,duplicated_bylinks:{ kind, url, title, createdAt }attachments:{ fileId, name, mime, size }, by referencepullRequests:{ kind, repo, number, sha, title, url, state }history: the activity log,{ at, actor { kind, id, via }, action, field, from, to }, the first 200 entries
What an export never contains
An export holds what the person who asked can see. Issues of a private team (and a private sub-team) they are not in are left out, and so are their projects, labels, their comments and relations: a parent or related issue in such a team is left out of the row that points to it. Owners and admins see every team, so their workspace export holds every team. The workspace export keeps checking, at every step, that its requester is still an owner or admin, and stops if not.
Every export is recorded in the audit log (tasks.export.issues, projects, initiatives, members, workspace,
download) with who, how many rows and which door (app, API, MCP). Each person can start 30 exports an hour; past that the
answer is 429. A workspace ZIP is stored in Venom's database only while it can be downloaded, never in a public place,
and the download needs a signed-in session of the person who started it.
Bring it back in
issues.csv imports through Settings → Import with the Linear source: its columns carry what that reader takes
(ID, Title, Description, Status, Priority, Labels, Assignee matched by address, Estimate, Due date,
Created, Completed, Canceled, Parent ID for sub-issues). Statuses and people are mapped in the dry run as for any
import, labels lose their group, and the formula apostrophe is removed.
Print an issue
The issue's ⋯ menu has Print…: the browser's print dialog opens on a clean page with the title, the properties, the description, the sub-issues and the comments with their full timestamps, and none of the app around it.
For agents
| Tool | What it does |
|---|---|
tasks_issue_export |
The CSV text of a list (the filters of tasks_issue_search), up to the caller's limit and about 1.5 MB of text; past that it says to narrow the filter or start a workspace export |
tasks_project_export, tasks_initiative_export |
The tables as CSV text |
tasks_member_export |
The members table (tasks:admin) |
tasks_export_create |
Start the workspace export (tasks:admin); the answer carries the export id |
tasks_export_get, tasks_export_list, tasks_export_cancel |
Status (asking also moves a running export on), the list, cancel or delete |
The tools run the routes above as the caller, so the role, the private-team rule, the audit row and the rate limit are the
same. A finished export answers with a downloadPath; the file itself is never put in a tool result, and the link only
works for the person who started the export, signed in.