Issue Tools

Ten tools covering the full issue lifecycle on the board: create, create in bulk, update, delete, search, list-mine, prepare project/issue context, pick up an issue end-to-end, and return the daily Today snapshot.

create-issue

Create a new issue in a project.

Parameters

Parameter	Type	Required	Description
project_id	integer	Yes	The project to create the issue in
title	string	Yes	Issue title (max 255 characters)
description	string	Yes	Issue description (max 65,535 characters)
lane_id	integer	No	Lane to place the issue in. Defaults to the first lane if omitted.
priority	integer	No	0 Highest, 1 High, 2 Medium (default), 3 Low, 4 Lowest
type	integer	No	0 Feature (default), 1 Bug, 2 Task
assignee_id	integer	No	User ID to assign. Use project members to find IDs.
sprint_id	integer	No	Sprint to add the issue to
epic_id	integer	No	Epic to link the issue to. Bidirectional with update-epic's issue_ids.
label_ids	array	No	Label IDs to attach (must belong to the same project, at most 500). Omit or pass [] for none.
blocked_by_ids	array	No	Issue IDs that block this issue (must be in the same project, at most 500)
blocks_ids	array	No	Issue IDs that this issue blocks (must be in the same project, at most 500)
estimated_minutes	integer	No	Estimated time in minutes (minimum 0)

Behavior

• If lane_id is omitted, the issue is placed in the project's first lane automatically
• The lane, sprint, and assignee are all validated to belong to the specified project
• Blocking relations must reference issues within the same project
• In the description, a plain @Full Name (or unique first name) of a project member becomes a real mention and notifies that member. Ambiguous or unknown names stay literal text, and names inside code blocks are never matched
• Returns the created issue with its generated key (e.g., KD-0042), ID, lane, assignee, sprint, blocking relations, and created_at/updated_at (ISO 8601)

Example Prompt

Create a high-priority bug in the kendo project: "Login form rejects valid email addresses". The form strips everything after a + in the address. Assign it to Jasper.

bulk-create-issues

Create up to 25 issues in one call. Use this instead of calling create-issue repeatedly when importing or generating a backlog — 250 issues becomes 10 calls rather than 250.

Parameters

Parameter	Type	Required	Description
project_id	integer	Yes	The project to create the issues in
lane_id	integer	No	Lane for every issue in the batch. Defaults to the project's first lane. One lane applies to the whole batch.
rows	array	Yes	1–25 issues to create. Each row: title (required, max 255), description (required, max 65,535), priority (optional), type (optional)

Behavior

• Every issue lands in the backlog. No sprint_id is accepted, so created issues appear on the Backlog tab, not the board. Assign them afterwards with assign-issues-to-sprint, or link them to an epic with update-epic.
• One lane for the batch. Per-row lanes aren't supported; move individual issues afterwards if they need different columns.
• A malformed row rejects the entire call. Because every row field is validated up front, one row missing a title — or carrying an out-of-range priority — means nothing is created. Fix the row and resend the whole batch; there is no risk of duplicates.
• A row that fails during creation does not stop the others. Infrastructure faults (a database error, a lock timeout) are reported per row in failed while its siblings still commit. The error text is fixed guidance, not a diagnosis — the underlying cause is recorded server-side rather than returned, so retry a failed row once and report it if it fails again.
• Batch-created issues are indistinguishable from individually created ones — same key format, ranks, watchers, audit entries and lane placement — because this tool reuses the same creation path as create-issue.
• @Full Name mentions in a description work exactly as they do on create-issue, and notify that member per issue.
• Only these fields are supported per row. Assignee, sprint, epic, labels, blocking relations, estimates and attachments are deliberately out of scope — set them afterwards with update-issue or the bulk-assign tools.

Response Fields

Field	Type	Description
created_count	integer	How many issues were created
failed_count	integer	How many rows failed
complete	boolean	true when every row was created
keys	array	Display keys of the created issues (e.g. KD-0142), in the order the rows were submitted
failed	array	Rows that failed. Each entry has index — the 0-based position in the rows array you sent — and error, a fixed guidance string. Resend only those rows, once; the others exist.

The response is deliberately thin: no descriptions are echoed back and no URLs are included, because mirroring create-issue's full payload across 25 rows would cost more tokens than the batching saves.

Rate Limits

A tenant may create up to 250 rows per minute (100 on the free plan). The batch is refused before anything is created, and the error names both the limit and when it resets. This is separate from the endpoint's own request limit — it exists because one bulk request carries 25× the writes of a single create.

Example Prompt

Take the twelve features from our roadmap doc and create them as issues in the kendo project backlog, with the two auth ones as Highest priority.

update-issue

Update an existing issue. Only provided fields are changed — omitted fields keep their current values.

Parameters

Parameter	Type	Required	Description
issue_id	integer	One of	The numeric ID of the issue (provide issue_id or issue_key, not both)
issue_key	string	One of	The issue key, e.g. "KD-0042" (provide issue_id or issue_key, not both)
title	string	No	New title (max 255 characters)
description	string	No	New description (max 65,535 characters)
lane_id	integer	No	Target lane ID to move the issue to (must be in the same project)
priority	integer	No	0 Highest, 1 High, 2 Medium, 3 Low, 4 Lowest
type	integer	No	0 Feature, 1 Bug, 2 Task
assignee_id	integer	No	User ID to assign, or null to unassign
sprint_id	integer	No	Sprint ID, or null to remove from sprint. For multiple issues at once, prefer assign-issues-to-sprint — one batched call instead of looping.
epic_id	integer	No	Epic ID, or null to unlink from epic. Bidirectional with update-epic's issue_ids. An issue that leaves an epic loses its links to that epic's Done when wishes.
blocked_by_ids	array	No	Replaces all "blocked by" relations (must be in the same project, at most 500). Omit to preserve, [] to clear.
blocks_ids	array	No	Replaces all "blocks" relations (must be in the same project, at most 500). Omit to preserve, [] to clear.
estimated_minutes	integer	No	Estimated time in minutes, or null to clear

Behavior

• Partial updates: only fields you include are modified, like the API's PATCH. The API's PUT requires all fields.
• Lane changes: pass lane_id to move an issue to a different lane on the board. The lane must belong to the same project.
• Blocking relations: omitting blocked_by_ids/blocks_ids preserves existing relations. Pass an empty array [] to explicitly clear them.
• Mentions: in the description, a plain @Full Name (or unique first name) of a project member becomes a real mention and notifies that member. Ambiguous or unknown names stay literal text.
• Pass null for assignee_id, sprint_id, or estimated_minutes to clear those values.
• title, description, lane_id, priority and type cannot be cleared: null on any of them is rejected. Omit a field to keep its value.
• Response includes created_at and updated_at (ISO 8601)

Example Prompt

Bump KD-42 to highest priority and assign it to Sarah

delete-issue

Permanently delete an issue and all its comments.

Destructive

This action cannot be undone. The issue and all associated comments are permanently removed.

Parameters

Parameter	Type	Required	Description
issue_id	integer	One of	The numeric ID of the issue (provide issue_id or issue_key, not both)
issue_key	string	One of	The issue key, e.g. "KD-0042" (provide issue_id or issue_key, not both)

Example Prompt

Delete the duplicate issue KD-38, it's the same as KD-42

search-issues

Search, filter or count issues within a project by text query and/or filters. This tool is read-only and safe to call repeatedly.

Parameters

Parameter	Type	Required	Description
project_id	integer	Yes	The project to search in
query	string	No	Text search across title, description, and issue key (max 200 characters)
lane_ids	integer[]	No	Issues in any of these lanes. All IDs must belong to the project
assignee_ids	integer[]	No	Issues assigned to any of these users. 0 means unassigned, so [0, 12] is unassigned or user 12
sprint_ids	integer[]	No	Issues in any of these sprints. 0 means no sprint (backlog). Other IDs must belong to the project
epic_ids	integer[]	No	Issues in any of these epics. 0 means no epic. Other IDs must belong to the project
priorities	integer[]	No	Issues with any of these priorities (0–4)
types	integer[]	No	Issues of any of these types (0 Feature, 1 Bug, 2 Task)
label_ids	array	No	Filter by label IDs — returns issues carrying ANY of these labels (OR). All IDs must belong to the project.
exclude_lane_ids	integer[]	No	Leave out issues in these lanes
exclude_assignee_ids	integer[]	No	Leave out issues assigned to these users. Unassigned issues stay unless 0 is in the list
exclude_sprint_ids	integer[]	No	Leave out issues in these sprints. Backlog issues stay unless 0 is in the list
exclude_epic_ids	integer[]	No	Leave out issues in these epics. Issues without an epic stay unless 0 is in the list, so [0] returns only issues that have an epic
exclude_label_ids	integer[]	No	Leave out issues carrying any of these labels
exclude_priorities	integer[]	No	Leave out issues with these priorities
exclude_types	integer[]	No	Leave out issues of these types
lane_id, assignee_id, sprint_id, epic_id, priority, type	integer	No	One value for the same filter; the tool adds it to the matching list, so assignee_id, sprint_id or epic_id of 0 means none, as 0 does in the list. lane_id must be at least 1
group_by	string	No	lane, assignee, sprint, epic, priority, type or label. Returns counts per value instead of issues
include_description	boolean	No	Default true. Pass false to leave description out of each issue
limit	integer	No	Max results (1–100, default 25)

Behavior

• Results are ordered by most recently updated
• All filters are optional and combine with AND logic
• The values inside one list combine with OR. When a value is in an include list and in the exclude list for the same field, the exclusion wins
• A lane, sprint, epic or label ID that does not belong to the project returns an error. Assignee IDs are not checked: a user with no issues in the project matches nothing
• Returns matching issues with key, title, lane, assignee, sprint, epic, labels, priority, and type
• Each result names its lane, assignee, sprint and epic next to their IDs (lane + lane_id, epic + epic_id, …), so the name answers the user and the ID feeds the next call
• Each result carries labels as {id, name} objects, plus label_ids
• Each result carries priority and type as numbers, plus priority_name (Highest … Lowest) and type_name (Feature, Bug, Task)
• Each result includes attachments_count — the number of files attached to the issue
• Each result includes created_at and updated_at (ISO 8601)
• Marked as #[IsReadOnly] and #[IsIdempotent] — safe to call multiple times with no side effects

Response Fields

Alongside the issues array, the response surfaces truncation metadata so the caller knows when the result set was capped:

Field	Type	Description
count	integer	Number of issues returned, never above limit. It is not the number of matches: read truncated, or pass group_by for the uncapped total
truncated	boolean	true when more issues matched than were returned; refine filters to see more matches
limit	integer	Server-side cap applied to this request
issues	object[]	Matching issues. description is left out when include_description is false

Counting with group_by

With group_by, the tool answers a question such as "open issues per lane, without Tasks" in one call, and no issue rows are returned. limit and include_description are ignored. The response carries project, project_id and:

Field	Type	Description
group_by	string	The field the counts are grouped on
total	integer	Number of issues that match the filters (not capped)
groups	object[]	{key, label, count} per value, highest count first. key is the lane, user, sprint, epic or label ID, or the priority / type number. key 0 is the group without a value for assignee, sprint, epic and label

With group_by: "label" an issue with several labels is counted in each of its label groups, so the group counts can add up to more than total.

Example Prompt

Show me all high-priority bugs assigned to Jasper in the kendo project

get-my-issues

Get issues assigned to you — your current work, tasks, and tickets. Excludes issues in the "done" lane (last lane per project).

Parameters

Parameter	Type	Required	Description
project_id	integer	No	Filter to a specific project (omit for all accessible projects)
priority	integer	No	Filter by priority (0–4)
type	integer	No	Filter by type (0 Feature, 1 Bug, 2 Task)
query	string	No	Text search across issue title and key (max 200 characters)

Behavior

• Returns only issues assigned to the authenticated user
• Excludes issues in the last lane (done) of each project
• Results ordered by most recently updated
• Each result includes created_at and updated_at (ISO 8601)
• Each result names its lane, project, sprint and epic next to their IDs (lane + lane_id, project + project_id, …), so the name answers the user and the ID feeds the next call
• Each result carries labels as {id, name} objects, plus label_ids
• Each result carries priority and type as numbers, plus priority_name (Highest … Lowest) and type_name (Feature, Bug, Task)
• Respects project accessibility — non-admin users only see issues in projects they can access
• Marked as #[IsReadOnly] and #[IsIdempotent]

Response Fields

Alongside the issues array, the response surfaces truncation metadata so the caller knows when the result set was capped at the server-side cap (currently 500):

Field	Type	Description
count	integer	Number of issues actually returned
truncated	boolean	true when the result set was capped at limit; refine filters to see more matches
limit	integer	Server-side cap applied to this request
issues	object[]	List of assigned issues

Example Prompts

What am I working on right now?

Show me my high-priority bugs across all projects

What tickets do I have in the kendo project?

prepare-project-context

Bundle the project's structural meta into one read: the project itself, all its lanes (board columns), all Planned and Active sprints (with active_sprint as a shortcut to the currently Active one), the project's labels, the project members (for assignee lookup), and the calling user. Use this as the gather step for any project-scoped flow — triage, board sync, branch creation, picking up an issue. Replaces separate reads of kendo://projects/{id}, kendo://projects/{id}/lanes, kendo://projects/{id}/sprints, and kendo://projects/{id}/members. For Completed sprints, use get-sprints with status=2.

Pair with prepare-issue-context

When you also need a specific issue's payload (comments, branch links, attachments), fire prepare-project-context and prepare-issue-context in parallel — same wall-clock as one round-trip. See Workflows → Prepare an issue.

Parameters

Parameter	Type	Required	Description
project_id	integer	yes	The numeric ID of the project

Behavior

• Marked as #[IsReadOnly] and #[IsIdempotent] — safe to call repeatedly with no side effects
• sprints contains only Planned (0) and Active (1) sprints, ordered by start date. Completed sprints are excluded — use get-sprints with status=2 for those
• active_sprint is a shortcut to the Active sprint from sprints, or null when none is Active
• members returns every user who reaches the project via team membership or direct membership, ordered by first name then last name
• members_count companion field lets you shortcut without iterating
• current_user is the authenticated MCP user — use this instead of guessing from git config user.email when deciding the assignee for a "start working on this" flow
• The full issue list is not included; for a single issue's payload (comments, branches, attachments), call prepare-issue-context — typically in parallel

Response Fields

Field	Type	Description
project	object	Project payload — the same fields as kendo://projects/{id}, including its archived state
lanes	object[]	Project lanes ordered by position (id, title, color, order, issues_count)
sprints	object[]	Planned and Active sprints ordered by start date (id, title, status, status_name, start, end, issues_count). Completed sprints excluded.
active_sprint	object | null	Shortcut to the currently Active sprint from sprints (id, title, status, status_name, start, end, issues_count), or null
labels	object[]	Project labels ordered by position (id, name, color, order) — use label.id as label_ids in search-issues or sync-issue-labels
members	object[]	Project members ordered by first_name, then last_name (id, name, email) — use member.id as assignee_id
members_count	integer	Total number of members returned
current_user	object	The authenticated MCP user (id, name, email)

Example Prompt

Pull together the project context for kendo: lanes, sprints, labels, members, and tell me who I am.

prepare-issue-context

Bundle a single issue's full payload into one read: the issue (with comments, branch links, time spent, attachments) and its linked epic. For project meta (lanes, active sprint, members, calling user), call prepare-project-context — typically in parallel with this tool.

Pair with start-work-on-issue

Call prepare-issue-context and prepare-project-context in parallel to gather, then start-work-on-issue to act (assign, move lane, link branch) in one idempotent write. See Workflows → Prepare an issue.

Parameters

Parameter	Type	Required	Description
issue_id	integer	One of	The numeric ID of the issue (provide issue_id or issue_key, not both)
issue_key	string	One of	The issue key, e.g. "KD-0042" (provide issue_id or issue_key, not both)

Behavior

• Marked as #[IsReadOnly] and #[IsIdempotent] — safe to call repeatedly with no side effects
• epic is null when the issue has no linked epic
• Project meta (lanes, active sprint, members, current user) is not included by design — use prepare-project-context for that, in parallel when you need both

Response Fields

Field	Type	Description
issue	object	Full issue payload — same shape as kendo://issues/{key} (id, key, title, description, priority, type, project, lane, assignee, sprint, comments — each with created_at/updated_at, blocked_by_ids, blocks_ids, labels — each {id, name, color}, epic_id, user_id, branch_links — each with created_at/updated_at, estimated_minutes, total_minutes_spent, attachments, created_at, updated_at)
epic	object | null	Epic linked to the issue (id, title, goal, description, status, status_name, color, color_name, start, end, checks — its What done looks like checks in order, each {id, text}, wishes — its Done when wishes in order, each {id, text, issues, checked_at, checked_by_id} with issues as {id, key}, created_at, updated_at), or null

Example Prompt

I want to start working on KD-42. Fire prepare-issue-context and prepare-project-context in parallel and tell me what I'm picking up.

start-work-on-issue

Pick up an issue in one idempotent call — assign it, move it to the target lane (e.g. "In Progress"), optionally add it to a sprint, and link a git branch. Re-running with the same values is a no-op. The GitHub repository is auto-resolved from the project's primary repo.

Pair with the prepare-context tools

Call prepare-issue-context and prepare-project-context in parallel to gather the issue payload and project meta, then start-work-on-issue to act — assign, move lane, link branch — in one write. See Workflows → Prepare an issue.

Parameters

Parameter	Type	Required	Description
issue_id	integer	One of	The numeric ID of the issue (provide issue_id or issue_key, not both)
issue_key	string	One of	The issue key, e.g. "KD-0042" (provide issue_id or issue_key, not both)
branch_name	string	Yes	The git branch name to link to the issue
assignee_id	integer	No	User ID to assign the issue to
lane_id	integer	No	Target lane ID to move the issue to (must be in the same project)
sprint_id	integer	No	Sprint ID. Omit to preserve existing; pass null to remove from sprint; pass an ID to set.
repo_id	integer	Conditional	Target GitHub repo ID. Omit on single-repo projects. On multi-repo projects, omit on the first attempt; if you get a candidate-listing error, retry with one of the listed repo IDs (look up IDs via the kendo://projects/{id}/github-repos resource) or ask the human to pick.

Behavior

• Idempotent — re-running with the same branch_name, assignee_id, lane_id, and sprint_id produces no further state change
• The branch is linked against the project's only GitHub repo when there is exactly one — no repo_id needed
• On multi-repo projects without repo_id, the tool returns a candidate-listing error rather than guessing — retry with the right repo_id
• Returns the updated issue (with lane, assignee, sprint, created_at/updated_at) and the branch-link record (existing on replay, otherwise newly created; carries status as the int, status_name beside it, and created_at/updated_at)

Example Prompt

Start work on KD-0042: assign it to me, move it to "In Progress", add it to the active sprint, and link branch KD-0042-fix-login-form.

kendo-today

Returns the user's active sprints, open issues grouped per project, and today's logged minutes as a single atomic snapshot. Prefer this over get-my-issues for daily-overview questions like "what's on my plate today".

Parameters

No parameters — operates on the authenticated user.

Response Fields

Field	Type	Description
activeSprints	object[]	Active sprints across accessible projects — each with id, title, projectId, projectName, myOpenIssuesCount, created_at, updated_at, projectAvatar
myIssues	object[]	User's open issues (excluding final-lane) — each with id, key, title, type (feature | bug | task), projectId, projectName, created_at, updated_at, projectAvatar
todayMinutes	integer	Total minutes logged by the user today across accessible projects

projectAvatar is the row's project avatar as absolute avif and webp URLs, the same object as the REST API's project avatar, or null when the project has none.

Example Prompt

What's on my plate today?

See Also

• Resources — Read issue data via kendo://issues/{id} and kendo://projects/{id}/issues
• Issues & Board Guide — Understanding the board, lanes, and blocking relationships
• Issues API — REST endpoints for issue management