Niches · Project & task management

Coding agent picks up a ticket, works it and reports back on the ticket

No company's docs show a full fit yet. The listings below do the job with a limit, or not at all.

0companies, full fit
17companies, fit with a limit
20companies carry the tag
21share of the category's work
12first-hand accounts

What counts as done

The agent reads the ticket with its description, comments and attachments in one call, posts at least a plan comment and a completion comment, and changes status to the done state, with each write visible in the tracker's activity history.

How agents ask for it: pick up the next ticket assigned to you, work on PROJ-123, comment your plan on the issue before you start, move it to done and summarise what changed

Does the job, with a limit 17

It does the job, but something in the docs or a user's account will get in an agent's way. The limit is named in each row.

ListingWhy
Asana BB
Asana, Inc.
get_task returns comments by default, get_attachments gives download URLs, add_comment and update_tasks complete it, but attachments are a second call, about 44 tools load at once and MCP grants carry no scopes. source
Basecamp B
37signals LLC
To-dos, comments and uploads are separate endpoints, so a ticket read takes several calls; completion and comment writes exist, and basecamp mcp --read-only and --domains narrow the tool set. source
ClickUp B
Mango Technologies, Inc. DBA ClickUp
Get Task, Get Task Comments and Download Task Attachment read a ticket and attachments, comment and status writes exist, but that is three calls and 65 tools load with no profile or read-only grant. source
monday.com BB
monday.com Ltd.
get_updates, get_assets and create_update cover reading comments and files and posting, and status is a column write, but reads span several calls, 60+ tools load, and acc-25 found read-only mode left only ids and names. source
ONES D
The Beautiful ONES Technology PTE. LTD.
get_issue_details, get_list_of_issue_comments, download_attachment, post_issue_comment and execute_issue_workflow cover the loop and scopes can be narrowed per category, but the read is three calls and 60+ tools load at once. source
OpenProject D
OpenProject GmbH
MCP reads a work package and its comments, posts comments and updates status, but no MCP tool reads attachments (APIv3 does), and the server is an Enterprise add-on, which pushed acc-54 to build its own. source
Plane B
Plane Software, Inc.
retrieve_by_identifier takes ENG-42, attachment read returns images and text inline, and comments and state updates are actions, but comments and attachments are separate calls; acc-16 runs this loop with AGENTS.md rules. source
Redmine E
Redmine project
GET /issues/<id>?include=attachments,journals returns description, comments and attachments in one call, and a PUT with notes and status_id comments and closes it, but the key carries its user's full permissions and there is no MCP server. source
Roma D
Milo Mode Inc.
Items can be delegated to a registered agent, get_item with includeFiles returns attached file text and image URLs, and post_chat_message posts on the item's thread, but every token reaches the whole workspace with no read-only grant, and 42 tools load. source
Shortcut B
Shortcut Software Company
The MCP has full story support including comments, and scopes split read, story-write and comment-write, but acc-34 found the MCP can upload files and not download them, so screenshots on a ticket can't be read. source
Taiga F
Taiga Cloud Services, S.L.
Issues and user stories have get, comment-by-edit, attachment list and status endpoints over REST, but reading description, comments (history) and attachments takes three calls, and the token is a 24-hour login with the account's full permissions. source
Taskade C
Taskcade Inc.
read_project returns the task tree with assignees and dates and complete_task closes a task, but no MCP tool reads or posts task comments and attachments aren't readable per task; the hosted server also needs a paid plan. source
Teamwork.com B
Teamwork Crew Limited
Task get, comment list and create, file get and download_file and complete_task cover the loop, and profile endpoints narrow tools, but comments and files are separate calls and acc-51 saw an EU site fail to connect from Codex. source
Todoist B
Todoist Inc.
fetch-object, find-comments, view-attachment, add-comments and complete-tasks cover the loop, but the read is several calls, 47 tools load at a fixed cost near 35,700 tokens, and the hosted server grants read and write only. source
Trello B
Trello, Inc.
A card read can include attachments and comment actions, attachment and comment endpoints exist, and moving a card to a Done list is a field change, but no official MCP server exists and an unattended agent relies on the key and token, which can be read-scoped. source
Wrike C
Wrike, Inc.
get_item_details returns status, assignees and attachments, get_item_comments reads up to 200 comments, and create_item_comment and update_items write back, but that is two reads, attachments come as metadata and links, and acc-52 saw OAuth token exchange fail in Claude Code. source
YouTrack F
JetBrains s.r.o.
get_issue, get_issue_comments, add_issue_comment and update_issue cover the loop with a URL returned, and ?tools= narrows 23 tools, but no MCP tool reads attachments, and acc-53 saw goose fail to initialise against YouTrack 2026.2. source

Doesn't do this job 1

Listings in the category that do something else, with the reason.

ListingWhy
Financial Cents MCP E
Financial Cents LLC
Financial Cents tracks accounting-firm work; manage_project_task has no comment field and no task comment tool appears, so a coding agent cannot post plan and completion comments on a ticket. source

What stops agents finishing

  • MCP server loads dozens of tools and fills the coding agent's context
  • attachments can be uploaded but not read back
  • no credential that limits a headless agent to its own tickets
  • OAuth connect fails in the coding client (redirect URI rejected, no dynamic client registration)

Around it

Before: project-management.find-work. After: code.open-pull-request, messaging.post-update.

Where this comes from

Weight 21 of 100 in Project & task management: 12 of 58 accounts; HN commenters running Jira, Linear, Plane, Redmine, Asana and monday boards as the queue for Claude Code or Codex, plus four who dropped tracker MCPs for a markdown file, a CLI or their own board because of token cost or broken servers. Four accounts are on Jira and Linear, which are not listed; they show the job, not a listing.

The same as data: /api/v1/use-cases/project-management.coding-agent-ticket-loop.json, or the MCP tool get_use_case with id project-management.coding-agent-ticket-loop. Other jobs in this category: /api/v1/niches?category=project-management.

Draft, read 2026-11-15 from public docs and first-hand accounts, not yet reviewed by a person. A mark says what the docs show, not what a run measured.

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.