Setting Work Item 456 to Done While Linking a GitHub Pull Request to Items 456 457 and 458

Why the Answer Is Correct

Two things must happen in one commit message: all three work items must be linked to the pull request, and item 456 must be moved to done. Because the code is in GitHub, the references have to use the Boards identifier form with the AB prefix, which is what lets GitHub and Azure Boards associate the pull request with items 456, 457, and 458. The resolution keyword Fixes applied to 456 is what performs the state transition to done, since it is one of the supported keywords that Azure Boards acts on. The vote was 50 for C and 25 for B, with 25 left uncast in the exported data. tratratra cited the Microsoft Learn page for linking from GitHub to Boards work items, which documents exactly the prefixed reference form used in option C.

Why the Other Options Are Wrong

Fixes #456, #457, #458 (B) uses bare numeric references, and Serezhka raised the correct objection that the proper notation in this flow should carry the AB prefix. Because the pull request lives in GitHub, the unprefixed numbers do not resolve to Boards work items, so the requirement to link the pull request to all three items is not satisfied, even though the keyword would still act on 456. Done #456, #457, #458 (A) fails twice: Done is not among the supported resolution keywords, and the references are again unprefixed. danisancho cited the auto-complete-work-items documentation for option A, but that page describes referencing work items in a pull request description, which is a different mechanism from state transition by commit message, so the citation does not rescue the keyword. The bare list #AB456, #AB457, #AB458 (D) contains no resolution keyword at all, so nothing sets 456 to done.

Community Comment Notes

The vote was unusually fragmented at 50 for C, 25 for B, and 25 with no recorded answer. The strongest critic of the community's choice was Christian_garcia_martin, who argued that C is wrong because items written as #A8457 and #A8458 will not be linked, suggesting they would not resolve, and concluded the intended answer was B. The counter is that ahmedMN explicitly addressed the apparent 7 versus 8 discrepancy and concluded the 8 is not a typo in the source material, favoring C, while megleg and Christian_garcia_martin both emphasized the #AB prefix requirement that only C provides. The community is split on whether the malformed digits are intentional, which is the only point of disagreement.