A sub-issue can have only one parent, so when --add-sub-issue is given
a candidate that already belongs to another parent the addSubIssue
mutation 422s with "Sub issue may only have one parent". Pass
replaceParent: true on this path so the take-over is silent.
The field gates whether the interactive survey offers the field as a
choice; rename so the name reflects intent and add a godoc.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The --type filter was concatenated into ImmutableKeywords in
SearchQueryBuild, which bypassed search.Query's quoting and let values
that contained quotes (or extra qualifiers) corrupt the final query.
Per babakks's suggestion in the review thread, model it as a regular
qualifier instead.
Add an IssueType field to Qualifiers tagged `qualifier:"type"` so it
shares a key with the existing Type field, and rework Qualifiers.Map
to honour the tag and concatenate values when multiple fields share
the same key. SearchQueryBuild now drops its bespoke type:X
formatting and just sets Qualifiers.IssueType, leaving the
keyword/qualifier escaping to pkg/search.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Pair --type with --remove-type so callers can clear an issue's type
without going through the interactive editor, mirroring the
--milestone / --remove-milestone and --parent / --remove-parent
patterns. The two type flags are mutually exclusive.
UpdateIssueIssueType now sends a null issueTypeId when the caller
passes an empty string, which is what the API requires to clear the
field. The orchestrator fires the mutation when either IssueTypeID is
non-empty or RemoveIssueType is set.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The post-creation Issues 2.0 mutations (issue type, parent, blocked-by,
blocking) ran sequentially in three separate apply* helpers. Replace
them with a single call to api.DeferredUpdateIssue, which fans the
mutations out in parallel and joins their errors. The new
newCreateDeferredOpts helper resolves the user-supplied refs to node
IDs (re-using the cached opts.issueTypeID from the interactive prompt)
and hands them to the orchestrator.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The --milestone / --remove-milestone pair has long been the
established pattern for value-takes-or-removes flag pairs in
gh issue edit. Bring --set-parent into line as --parent / --remove-parent.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The interactive Parent prompt was a free-text input with no
candidate-listing UX, an oversight from the initial Issues 2.0
landing. Sub-issues, blocked-by, and blocking already live as bare
flag fields outside of Editable for the same reason; bring Parent
into line.
editRun now reads opts.SetParent / opts.RemoveParent directly when
constructing DeferredUpdateIssueOptions, and the survey machinery
no longer sees Parent at all.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The Issues 2.0 mutations (issue type, parent set/remove, sub-issues,
blocked-by, blocking) are deferred until after the main UpdateIssue
because they target IDs that the standard mutation does not handle.
Move them behind a shared api.DeferredUpdateIssue orchestrator that
fans them out in parallel and joins all errors so a single failure
does not abort the rest.
editRun no longer carries its own applyEditParent / applyEditSubIssues
/ applyEditRelationships helpers; the per-issue goroutine resolves
refs to node IDs via a small deferredUpdateIssueOptions builder, then
hands the populated DeferredUpdateIssueOptions to api.DeferredUpdateIssue.
Also moves ResolveIssueRef and ResolveIssueTypeName from the deleted
resolve.go into lookup.go.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Let the API return its own "unsupported" error rather than gating the
relationship mutations behind a client-side IssueRelationships check.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Non-interactive mode resolves the type once before the loop instead of
every iteration. Interactive mode still resolves inside the loop after
the survey populates IssueType.Value. Both paths share a small
lookupIssueTypeID helper.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
A sub-issue can have only one parent, so applying --add-sub-issue
across multiple parent args is ambiguous.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Add issue-type, parent, sub-issues, sub-issues-completed, blocked-by,
and blocking lines to the raw issue preview. Empty values still print
to keep line counts stable for head|grep workflows.
Pull the issue-ref formatting into a small set of helpers so the human
and machine renderers share a single owner/repo#N source of truth.
formatLinkedIssueRef no longer takes baseRepo: callers in this package
always have repository.nameWithOwner on the LinkedIssue, so the
disambiguation between same-repo and cross-repo references is no
longer needed and the resulting refs are unambiguous.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Relationship mutations (parent, sub-issue, blocked-by, blocking) run
against baseRepo's host with node IDs that must come from that same
host. Catch a different-host ref up front with a clear error instead
of letting the mutation fail with a confusing node-ID error.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
prShared.FetchOptions already fetches RepoIssueTypes when IssueType is
edited, but only kept the names. editRun then called RepoIssueTypes a
second time via ResolveIssueTypeName.
Have FetchOptions store the name->ID map on Editable, and look the ID
up directly in editRun. The lookup now also lives inside the per-issue
loop, which fixes a bug where the interactive type prompt's chosen
value was set after the upfront resolve, sending an empty issueTypeId
on the mutation.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Two values per relationship verifies each ref is resolved and applied
independently rather than masking a single-iteration bug.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Same code path, same scope. Test_createRun's httpStubs signature picks
up *testing.T to match the V2 cases that need it for input assertions.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The five Issues 2.0 cases (type, parent by number, parent by URL,
blocked-by, blocking) live as a separate parallel test today. They
exercise the same NewCmdCreate code path as the existing cases, so
fold them into TestNewCmdCreate and extend the assertions to cover
IssueType, Parent, BlockedBy, and Blocking.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The subIssues, blockedBy, and blocking JSON output is currently shaped
as a flat array, which silently truncates when there are more entries
than the GraphQL fragment fetches. That's misleading once a user
crosses the page size, so this switches each connection to a
{nodes, totalCount} object so consumers can see when there's more.
While confirming page sizes, the GitHub limits turn out to be:
- sub-issues: up to 100 per parent
https://docs.github.com/en/issues/tracking-your-work-with-issues/using-issues/adding-sub-issues
- blocked-by / blocking: up to 50 per relationship type
https://github.blog/changelog/2025-08-21-dependencies-on-issues/
So subIssues moves to first:100 to fetch the full set; blockedBy and
blocking stay at first:50, which already covers their cap.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The AddBlockedByPayload and RemoveBlockedByPayload types expose
the result as 'issue', not 'blockedIssue'. Found during live spec
testing against github.com — the mutations returned empty responses.
Updated mutation queries and corresponding test fixtures.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- Restore // TODO projectsV1Deprecation comment above
TestProjectsV1Deprecation (was displaced when new tests were
inserted at that location)
- Add 'relationships unsupported on GHES' test case to edit command
using DisabledDetectorMock (parity with create's GHES tests)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- Fix PR edit regression: gate Type and Parent behind Allowed bool
in FieldsToEditSurvey, matching the Reviewers.Allowed pattern.
Only issue edit sets Allowed=true; PR edit won't show these fields.
- Add missing RemoveBlocking assertion in flag parsing tests
- Quote issue type names containing spaces in search queries
(type:"Bug Report" instead of type:Bug Report)
- Remove duplicate TODO comment in view.go
- Avoid double RepoIssueTypes API call in interactive create:
cache the resolved ID from the picker, skip re-resolution
- Rename applyIssueTypes → applyIssueType (singular)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Completes the symmetry of relationship flags:
- --add-blocked-by / --remove-blocked-by
- --add-blocking / --remove-blocking
The --remove-blocking flag swaps API args (same as --add-blocking):
calls RemoveBlockedBy(issueId=OTHER, blockingIssueId=THIS).
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Add flag parsing and behavior tests for gh issue list --type:
- TestNewCmdList/type_flag: verifies opts.IssueType is set
- Test_issueList/with_issue_type: verifies search path is forced
and query includes type:Bug qualifier
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Critical fixes:
- GHES data-flow regression: blockedBy/blocking fields now
conditionally added to view lookupFields only when
IssueRelationshipsSupported is true (GHES 3.19+). Previously
would break gh issue view on GHES 3.17-3.18.
- State line separator: restore original bullet (•) to avoid
breaking downstream parsers. Issue type prefix uses middle dot (·).
Optimizations:
- Batch edit --type: resolve issueTypeID once before the loop
instead of per-goroutine (eliminates N-1 redundant API calls)
- Parent removal: include id in parent GraphQL fragment, use
issue.Parent.ID directly instead of extra IssueNodeID lookup
Nit fixes:
- Fix formatLinkedIssueRef godoc to match actual behavior
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Address code review findings:
- Extract resolveIssueRef into shared.ResolveIssueRef (was duplicated
between create.go and edit.go)
- Extract issue type name→ID resolution into shared.ResolveIssueTypeName
(was duplicated between create applyIssueTypes and edit applyEditIssueType)
- Fix double import of issue/shared in view.go
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Filter issues by type using the search API path. The --type flag
appends a type: qualifier to the search query, forcing the search
path (same as --label and --milestone).
Updated FilterOptions with IssueType field, IsDefault(), and
SearchQueryBuild() to include the type qualifier.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
New flags for issue edit:
- --type: set issue type by name
- --set-parent / --remove-parent: set or remove parent issue
(mutually exclusive via cmdutil.MutuallyExclusive)
- --add-sub-issue / --remove-sub-issue: manage sub-issues
- --add-blocked-by / --remove-blocked-by: manage blocked-by relationships
- --add-blocking: add blocking relationships (swaps API args)
Interactive mode: Type and Parent added to the field picker survey.
FetchOptions loads issue types when Type is selected.
Editable struct: added IssueType and Parent fields with Dirty(),
Clone(), FieldsToEditSurvey, and EditFieldsSurvey support.
GHES: relationships gated behind IssueRelationshipsSupported.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Post-creation mutations for Issues 2.0 fields:
- --type: resolve type name to ID via RepoIssueTypes, then
updateIssueIssueType mutation
- --parent: resolve issue ref (number or URL), then addSubIssue
mutation (supports cross-repo URLs)
- --blocked-by: resolve refs, then addBlockedBy mutations
- --blocking: resolve refs, then addBlockedBy with swapped args
Interactive mode: type picker when repo has issue types configured.
GHES: relationships gated behind IssueRelationshipsSupported feature
detection (3.19+). Types and sub-issues need no detection (GA 3.17+).
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Display new issue metadata in TTY view:
- Issue type on state line (gray, before Open/Closed)
- Type, Parent, Blocked by, Blocking metadata rows
- Sub-issues section with completion progress (X/Y, Z%)
- Cross-repo references show full owner/repo#N format
All new fields included in defaultFields and JSON export.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Two bugs introduced in #13009 found during acceptance testing:
1. `pr create --assignee @copilot` sent the literal `@copilot` to the
API because NewIssueState only ran MeReplacer on assignees, not
CopilotReplacer. Fixed by switching to SpecialAssigneeReplacer (which
handles both @me and @copilot) like issue create already does.
2. The replaceActorsForAssignable mutation requires the [bot] suffix on
bot actor logins (e.g. `copilot-swe-agent[bot]`), unlike
requestReviewsByLogin which has a separate botLogins field. Added the
suffix in ReplaceActorsForAssignableByLogin for CopilotAssigneeLogin.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Aligns the feature detector field name with the downstream
ApiActorsSupported flag introduced in the previous commit, so the
signal has one consistent name from detection through to consumption.
Also consolidates leftover TODO tags (actorIsAssignableCleanup,
requestReviewsByLoginCleanup) under the single // TODO ApiActorsSupported
tag so there's exactly one thing to grep for.
Pure rename with no logic changes.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The CLI had two per-entity flags (ActorAssignees on EditableAssignees and
IssueMetadataState, ActorReviewers on IssueMetadataState) threaded through
different layers of the stack to distinguish github.com from GHES. Both
flags were always set from the same source (issueFeatures.ActorIsAssignable)
and never had different values, but they were carried independently on
different structs. This led to a confusing asymmetry where:
- EditableAssignees had ActorAssignees but EditableReviewers had nothing
- The PR edit flow piggybacked on editable.Assignees.ActorAssignees to
make reviewer mutation decisions, which was misleading
- RepoMetadataInput only had ActorAssignees with no reviewer equivalent
This commit replaces all per-entity flags with a single ApiActorsSupported
bool hoisted to the shared level on Editable, IssueMetadataState, and
RepoMetadataInput. Both assignees and reviewers now key off the same signal.
Every branch site is marked with // TODO ApiActorsSupported so we can grep
for cleanup sites when GHES eventually supports the actor-based mutations
(replaceActorsForAssignable, requestReviewsByLogin).
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The inline replaceSpecialAssigneeNames closures in AssigneeIds and
AssigneeLogins were duplicated. Extract them into an exported
SpecialAssigneeReplacer type that consolidates MeReplacer and
CopilotReplacer into a single ReplaceSlice call, parameterised by
actorAssignees and copilotUseLogin.
Adopt the new type in the issue create flow as well, replacing the
manual MeReplacer + conditional CopilotReplacer sequence.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Wire up MultiSelectWithSearch for assignees in MetadataSurvey, replacing
the static MultiSelect that required bulk fetching all assignable actors.
This applies to both gh pr create and gh issue create interactive flows
when selecting assignees via the 'Add metadata' prompt.
Changes:
- Add assigneeSearchFunc parameter to MetadataSurvey
- Skip assignee bulk fetch when search func is available
- New SearchRepoAssignableActors API function for repo-level search
(create flows have no issue/PR node ID yet)
- New RepoAssigneeSearchFunc in shared editable.go
- Refactor actorsToSearchResult helper shared by both search functions
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Add AssigneeSearchFunc to gh issue edit interactive flow, matching
the pattern already used in gh pr edit. This eliminates the bulk
RepositoryAssignableActors fetch for interactive assignee selection,
using dynamic SuggestedAssignableActors search instead.
Also clean up pr edit assigneeSearchFunc signature to remove the
unused editable parameter (no longer needed after removing the
actor accumulation hack).
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
When ActorAssignees is true (github.com), the --add-assignee and
--remove-assignee flag flows now pass logins directly to
ReplaceActorsForAssignableByLogin instead of bulk fetching all
assignable actors and resolving logins to node IDs.
Changes:
- New AssigneeLogins() method on Editable that computes the final
login set (defaults + add - remove) without ID resolution
- UpdateIssue: call AssigneeLogins + ByLogin when ActorAssignees is true
- EditableOptionsFetch: skip assignee bulk fetch for flag flows on
github.com (only fetch on GHES where ID resolution is needed)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
When ActorAssignees is true (github.com), pass assignee logins directly
to the ReplaceActorsForAssignable mutation instead of resolving logins
to node IDs. This eliminates the need to bulk fetch all assignable
users/actors and fixes a bug where providing assignees via CLI flag
and then interactively adding metadata would fail with 'not found'
because the cached MetadataResult had no assignee data.
Changes:
- Set state.ActorAssignees = true in pr create (was missing)
- AddMetadataToIssueParams: pass assigneeLogins when ActorAssignees
is true, skip fetch and ID resolution entirely
- CreatePullRequest/IssueCreate: call ReplaceActorsForAssignableByLogin
after creation to assign via logins
- Consolidate replaceActorsForAssignable mutation into api/ package
(ReplaceActorsForAssignableByLogin + ReplaceActorsForAssignableByID)
- Remove duplicate replaceActorAssigneesForEditable from editable_http.go
- Add TODO replaceActorsByLoginCleanup markers on edit paths
Fixescli/cli#13000
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Only the destination repo ID is needed for issue transfer. Switch from
GitHubRepo to IssueRepoInfo to use minimal fields appropriate for issues.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Update test mocks to match the renamed GraphQL query used by
IssueRepoInfo, and switch to StubIssueRepoInfoResponse helper.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Switch issue create from GitHubRepo to IssueRepoInfo so that
gh issue create works with fine-grained PATs that only have
Issues:Write and Metadata:Read permissions.
Fixescli/cli#12798
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Apply DisplayName() to author and assignee display in issue view,
consistent with the pr view changes.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
The stateReason field was added in GHES ~3.4, which is far older than
the earliest supported GHES version (3.14). The feature detection and
conditional inclusion of stateReason is therefore unnecessary.
This removes:
- StateReason field from IssueFeatures struct
- GHES introspection query in IssueFeatures() (only ActorIsAssignable
remains, which is always false on GHES)
- Conditional stateReason field inclusion in issue list
- Feature detection guard in issue close
- Feature detection guard in FindIssueOrPR
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
