[muqsitnawaz]

What was wrong

Agents kept filing new issues into milestones whose target date had already passed. There were two causes:

1. The default milestone ignored the date. next_open_milestone picked the earliest targetDate still under 100% and never compared it to today. Once a milestone went overdue at less than 100%, every plain linear create in that project landed in it, labeled "next for ". On v0.33.0 against our live workspace, a default create picked a milestone whose target was 2026-10-23.
2. An explicit milestone was not checked. create --milestone, update --milestone and goals --milestone accepted a milestone whose target date had passed. Only --backlog checked the date.

What changes

• milestone_passed(m, today): a milestone whose target date is before the team's today. A milestone due today still takes issues.
• The default (next_open_milestone, used by create and goals) skips passed milestones.
• create, update and goals refuse a passed milestone with exit 1. The refusal names the milestone and its date, lists the project's milestones still open, and names linear milestones set-target-date for work that really belongs to it. Bulk --from-file rows get a one-line form.
• When no milestone is still open, the "no milestone" refusal lists only the open milestones.
• The CreateRequiresMilestoneTest fixture used hard-coded September 2026 targets, which are now past. It now uses dates relative to today, as _CreateWorld does.

New surface: milestone_passed and passed_milestone_error; resolve_milestone_id gains open_on. next_open_milestone now takes today. No new flags or config keys.

Run evidence

Captured run of this branch (831e847) against our live Linear workspace on 2026-11-03T17:19Z (org sign-in required, because it names internal milestones): https://share.getrush.ai/muqsit/linear-cli-no-past-milestone-run

What it shows:

• A default create now picks the next milestone due on or after today (target 2026-11-05) instead of the one that targeted 2026-10-23.
• linear create … --milestone <milestone that targeted 2026-10-23> prints the refusal and exits 1. The refusal lists the open milestones and the set-target-date command.
• linear update <issue> --milestone <milestone that targeted 2026-10-26> prints the same refusal and exits 1. Reading the issue afterwards showed no milestone was written.
• python3 -m unittest test_linear: Ran 288 tests … OK (skipped=4).

Goals: _goal_milestone defaults to the next open milestone, and an explicit passed milestone exits 1 with the same refusal.

🤖 Generated with Claude Code

[felicitycos1]

how it works sorry I don't understand

[muqsitnawaz]

Independent review (non-author code-reviewer subagent, read-only, 831e847): no blocking findings. It walked every writer of projectMilestoneId, covering create single and bulk, update single and bulk, goals set and goals todo, and checked the team-timezone "today", the resolve_milestone_id(open_on) error paths, the backlog ordering, and public-repo hygiene. Two non-blocking items were fixed in f07c1e7:

1. goals todo copied a passed Daily-goal milestone onto the new to-do. It now takes the next open milestone.
2. There were no tests for the update/goals refusals or open_on. Five are added, and all 10 new tests fail against v0.33.0.

python3 -m unittest test_linear: Ran 293, OK (skipped=4). CI: unittest and CodeQL pass.

[muqsitnawaz]

Verdict: APPROVE (independent non-author code-reviewer subagent). Its words: "READY TO MERGE after you decide on item 1 (a one-line fix) and add the missing update/goals tests from item 2." Both are done in f07c1e7, and CI passes on that head.

[felicitycos1]

Hi how it works

[felicitycos1]

okay if I understand I have to start from item one can I use any story. thanks Le mar. 3 nov. 2026, 17:36, Muqsit ***@***.***> a écrit :

…

*muqsitnawaz* left a comment (phnx-labs/linear-cli#79) <#79 (comment)> Verdict: *APPROVE* (independent non-author code-reviewer subagent). Its words: "READY TO MERGE after you decide on item 1 (a one-line fix) and add the missing update/goals tests from item 2." Both are done in f07c1e7 <f07c1e7>, and CI passes on that head. — Reply to this email directly, view it on GitHub <#79?email_source=notifications&email_token=BBOGAXEEDXN7P3IP2PX47435XIJRTA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTMNRRGEYDONZXGUYKM4TFMFZW63VHMNXW23LFNZ2KKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#issuecomment-6611077750>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BBOGAXD4JD27I4XGNRIZJLD5XIJRTAVCNFSNUABGKJSXA33TNF2G64TZHMYTEMRQGQZDMNBXGU5US43TOVSTWNRRHA3TMNZZGE4THILWAI> . Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/notifications/mobile/ios/BBOGAXFBUIVXRXSZOLRFBB35XIJRTA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTMNRRGEYDONZXGUYKM4TFMFZW63VHMNXW23LFNZ2KKZLWMVXHJKTGN5XXIZLSL5UW64Y> and Android <https://github.com/notifications/mobile/android/BBOGAXE5DP7BPKH5GQPDCT35XIJRTA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTMNRRGEYDONZXGUYKM4TFMFZW63VHMNXW23LFNZ2KKZLWMVXHJLTGN5XXIZLSL5QW4ZDSN5UWI>. Download it today! You are receiving this because you commented.Message ID: ***@***.***>