Which milestone did that issue join?

A client says the issue is in the release, or says it was never filed there. Tonight, three proofs, in a repository the agency or the client can open. Name the milestone that issue joined: the milestone number, the title, and the state. If the issue has no milestone, write that. Show a different issue whose number, or whose title, is not this one. Show another client's repository that does not list this milestone. If any of the three needs a laptop the client cannot open, you have not named the milestone. You have a title and a guess.

This is the milestone on one issue. It is not the label. Which label did that issue carry? is the name, the color, and the description. A label can be bug while the milestone number is empty, and a milestone can be open while the issue carries no label. It is not the release. Which release did that tag name? is the release id, the tag, and the draft flag. A tag can point at a release while the milestone title is a different string. It is not the branch. Which branch did that pull target? is the pull number, the base ref, and the head ref. A pull can be open while the milestone it mentions is a different object. It is not the quote a customer accepted, and it is not the run a thread started. Those are different objects, and they are not this number.

None of what follows is a promise about this client's date, and none of it is legal advice. The fields below are one host's current description of a milestone, and the list, the create, the get, and the update all print the same sample. The counts below that are a study of issues in other people's AI repositories. A closed count on an example is not a census of your client's issues. A paper that never opened this repository cannot name the milestone on the issue you opened tonight. They do tell you which numbers people mix up when they try to sound precise about a due date they have not opened.

GitHub's list response is an array of one milestone, and that one object is also what create, get, and update return. The object prints id 1002604, number 1, state open, title v1.0, and description Tracking milestone for version 1.0. The id is 7 digits. The number is 1 digit. 1 is not 1002604. The title is 4 characters. The description is 34 characters. 34 is not 4. open_issues is 4. closed_issues is 8. 4 plus 8 is 12. 12 is not 1, and 12 is not 1002604. The page does not print 12. Do not treat the sum as a field. The html address ends in v1.0. The API address ends in 1. The title and the number are both on the object, and they are not the same token. creator.login is octocat, 7 characters, creator.id is 1, and site_admin is false. 1 as a user id is not milestone number 1 just because both digits match. node_id is MDk6TWlsZXN0b25lMTAwMjYwNA==, 28 characters. 28 is not 7 (GitHub, REST API endpoints for milestones).

The dates on that same object do not agree with the state. created_at is 2011-04-10T20:09:31Z. due_on is 2012-10-09T23:39:01Z. closed_at is 2013-02-12T13:22:01Z. updated_at is 2014-03-03T18:58:10Z. Four timestamps, four years, and the state is still open. due_on minus created_at is 47359770 seconds. That is 548 days and 12570 seconds. 12570 seconds is 3 hours, 29 minutes, and 30 seconds. The gap is not 548 days. Do not drop the remainder until the page does. closed_at minus due_on is 10849380 seconds. That is 125 days and 49380 seconds, which is 13 hours and 43 minutes. A closed_at after the due date, on an object whose state is open, is the sample. It is not a reason to tell the client the milestone was closed. updated_at is later than closed_at. An update is not a close. The request body that create and update both send is title v1.0, state open, the same 34-character description, and due_on 2012-10-09T23:39:01Z. The request does not send open_issues, and it does not send closed_at. The response still prints 4, 8, and a closed_at in 2013. A create response that already has eight closed issues is a canned object. It is not the milestone you just made.

The query values and the stored values are not one list. List filters state by open, closed, or all. Three names. The stored state, and the state you may send on create, is open or closed. Two names. all is a filter. It is not a state you can read off the object. Default state for the list is open. Default sort is due_on, and the other sort is completeness. Two sort names. The page does not print a formula for completeness. 8 divided by 12 is 2/3. The page does not print that fraction, and it does not print 66.7%. Do not invent the formula because the division is available. direction is asc or desc, default asc. per_page default is 30, maximum 100. page default is 1. The list example returns 1 milestone. 1 is not 30, and 1 is not 100. Create returns 201. Get and update return 200. Delete returns 204 and an empty body. An empty body is not the object with id 1002604. The version header on the examples is 2026-03-10. That string is 10 characters. It is not the integer 1002604.

The study people reach for when they want a closed percentage did not open this milestone. Zhou Yang, Chenyu Wang, Jieke Shi, Thong Hoang, Pavneet Kochhar, Qinghua Lu, Zhenchang Xing, and David Lo study issues in open-source AI repositories. They collect 576 repositories and 24953 issues. Of the repositories, 122 have no issues, printed as 21.18%. 122 divided by 576 is 0.2118, which rounds to 21.18%. That quotient matches the print. 576 minus 122 is 454. The page does not print 454. Do not invent an average by dividing 24953 by a count the page left out. They say 8102 issues are open, printed as 32.47%, and 16851 are closed, printed as 67.53%. 8102 plus 16851 is 24953. The addition matches. 16851 divided by 24953 is 0.6753, which is 67.53% to two decimals and 67.5% to one decimal. The abstract prints 67.5%. The results print 67.53%. Those two prints are the same quotient rounded differently. Neither print is closed_issues 8, and neither is state open (Yang et al., What Do Users Ask in Open-Source AI Repositories?).

The days do not survive the conclusion. The abstract says half of the issues resolve within four days. The results say the median addressing time is four days, the average is 46.94 days, and the standard deviation is 135.34 days. 46.94 is not 4. 135.34 is not 46.94. The conclusion says half of them closed within five days. 5 is not 4. A median of four days is not a median of five days. The longest addressing time is described as over two years. Two years is not 46.94 days, and the page does not print the day count it would take to reach two years. 40% of closed issues are called self-closed. 40% of 16851 is 6740.4, which is not a whole issue. The page does not print 6740.4. Do not round it. Ignored issues are 11.79% of all issues. The page does not print the count that percent would make. 11.79 is not 67.53, and it is not the sample's closed_issues 8.

The taxonomy is a sample, and the sample is not the corpus. They draw 384 issues and sort them into 13 categories. The thirteen counts are 28, 75, 10, 42, 89, 35, 13, 16, 50, 5, 4, 15, and 2. Those thirteen numbers add to 384. The addition matches the sample. It does not match 24953. Runtime error is 89 issues, printed 23.18%. Unclear instructions is 75, printed 19.53%. 89 divided by 384 is 0.2318, which rounds to 23.18%. 75 divided by 384 is 0.1953, which rounds to 19.53%. Those two quotients match the table. 89 plus 75 is 164. 164 divided by 384 is 0.427. The page does not print 42.7%. The two most common categories are not the whole sample. 89 is not 24953, and 13 categories are not milestone number 1. The missing-content subcounts are 10, 9, and 9, which add to 28. The unclear-instruction subcounts are 56, 12, and 7, which add to 75. Those inner additions match. They are still counts of other people's issues. The paper's own API footnote names version 2022-11-28. That string is also 10 characters. It is not 2026-03-10.

Do the three proofs on one real milestone, in the repository that actually holds it. First, the binding. Copy the milestone number, the title, the state, the due_on, the open_issues, and the closed_issues. If state is open and closed_at is filled, write both. Do not rewrite open as closed because a timestamp exists. If the issue's milestone field is empty, write empty. Do not paste v1.0 onto it because a sample uses that title. Do not paste 4 and 8 because a sample prints them. Second, the other issue. A different issue, with a different number, or a milestone whose title is not this one, is not this milestone. A paper's closed percentage is not that second issue. If you cannot show an issue that is not on this milestone, you have a counter, not a record. Third, the other repository. A list that belongs to another client does not contain this number. If the only tracker is a spreadsheet on a laptop the client cannot open, write that. A repository you cannot name is the finding.

Three moves feel like the proof and are a second problem. Closing the milestone, or moving the issue, because the date was embarrassing, before you have copied the number. A state change does not tell you which milestone you were looking at. Quoting 67.5% because it looks precise, while the milestone you were asked about is one object, and the 24953 issues that produced that percent had not opened this repository. Asking a model which title it was. The model will pick a plausible string. It will not open the milestone. A label you have not kept, while the client is arguing about the word on the issue, is which label did that issue carry?. Opening that screen does not name the milestone number.

The path is ordinary when an operator can open one client's tracker without opening the next client's tracker as the way to prove it. Cove consolidates an operator portal, a content system, and a marketing platform into one operator experience. The shape that matters here is one client's systems, distinct from the next client's. That case is not a milestone ledger. Nothing in that sentence is a headcount, a duration, or a result figure. If the only control on the screen changes every client's issues together, you do not have this proof. You have a totals page.

There is a version of this afternoon where taking the work is the wrong purchase. The work is not sold under the agency's name. The client asked to meet the builders, and a hidden bench is the thing they said they did not buy. The room only has a slide, and the repository is not one you can open. The buyer only wants a template. Or the buyer wants a legal opinion about what a due date promises. Those are not this proof. The terms on a white label software development partner page are the arrangement itself: the work stays in the client's repositories, under the agency's name. The page is the offer. It does not name this milestone, and it does not turn 67.53% into a state.

When the issue is real and the repository cannot name the milestone, the missing object is specific. A read, in a repository the agency or the client can open, that shows one number, one title, and one state, shows a different issue that is not that milestone, and shows another client's repository without it. A white label software development partner that expects this will have put that read in the client's organization before the first issue, because a number that lives on a laptop the client cannot open cannot be defended by the person whose name is on the invoice. One that can only describe the milestone by forwarding a title is telling you who holds the tracker. If the next slice is one capability, an AI integration sprint is the shape where the acceptance sentence can include the three proofs. If you do not yet know which trackers the current engagements already use, a two-week AI readiness assessment maps them. The findings are yours. They are not a promise that a paper's percentage applies, and they are not 1002604.

The quote a customer accepted, and the run a thread started, are different objects: which quote did that customer accept? and which run did that thread start?. Do not record either on this page. A quote id is not a milestone number. A run id is not a title. The comment a review left is a different object: which comment did that review leave?. Do not record it here. A line number is not a milestone title. The team a member joined is a different object: which team did that member join?. Do not record it here. A team slug is not a milestone number.

Before you tell them the issue is in the release, do the three proofs and write the state. Name the number, the title, and the state, or write that the issue has no milestone. Show a different issue that is not that milestone. Show another repository that does not hold it. Do not close the milestone to avoid reading the state, and do not ask a model which of the sample titles it was. A paper can tell you how often other repositories' issues were closed, including the conclusion that says five days where the median is four. The binding is the milestone you can defend.