Get the next task from the queue based on priority and dependencies

Claim a task to work on it

Mark a task as completed. This will trigger validation questions based on project knowledge. If validation questions are returned, you must answer them by calling this tool again with the answer in the message parameter.

Create a new task

Update task details

Get detailed information about a task including dependencies and notes

List all epics for a project

Get epic details with all tasks

Create a new epic

Update epic details

Add a working note to a task (todo, bug, edge case, optimization, plan, or information). Plan notes auto-resolve previous plans. Priority defaults to informational for plan/information types.

Mark a working note as resolved

Escalate a working note to a new task

Get all working notes for a task

Sync an implementation plan to the currently claimed task as a working note. Uses the current task context automatically - no task_id needed. Previous plan notes are auto-resolved. Call this after exiting plan mode.

Add a blocking dependency between tasks (blocked_task waits for blocking_task to complete)

Remove a blocking dependency between tasks

Search project knowledge base (architecture, best practices, specs)

Add a new knowledge entry to the project

Post an activity event to track what the agent is doing. You MUST call this proactively throughout your work — do not wait to be asked. Specifically:

• Post progress after completing each significant milestone (e.g., "Database schema done, moving to API endpoints")

• Post file_created or file_modified when you create or change key files

• Post decision when you make architectural or technical choices (include rationale)

• Post commit after git commits, pr_created after creating PRs

• Post blocked if you cannot proceed, error if something goes wrong

• Post warning for things other agents or the team should know about The project_id and task_id are auto-filled from the currently claimed task if omitted.

Get recent activity events for a project or task

Get shared context for a project (contracts, utilities, decisions, config)

Set a shared context item so other agents can see your work. You SHOULD call this when:

• You create or modify a shared interface/type → category: "contracts"

• You create a reusable utility function → category: "utilities"

• You make a technical/architectural decision → category: "decisions" (include rationale)

• You add or change environment variables or config → category: "config" This enables parallel agents to stay coordinated without reading each other's code.

Delete a shared context item

Fetch the overview, manifest, and epic metadata from the PlanCreator wiki for a given project

List tasks from the PlanCreator wiki along with metadata to understand available plan steps

Retrieve the full markdown content and metadata for a specific plan task from the PlanCreator wiki

Search plan tasks in PlanCreator by title, description, tags, or content to find relevant context