maintain-tasks

Session continuity preserves task progress and intent across context compaction by using tasks as persistent working memory. This ensures a project's state and next steps remain immediately clear for multi-step work, even when the AI's context window is compacted.

Subagent coordination assigns tasks to specific subagents using a clear prefix convention in the task subject. This explicit agent assignment ensures clear responsibility delegation and prevents overlap when managing complex, multi-step project workflows.

Dependency management defines sequential or parallel task flows using `blockedBy` and `blocks` relationships. This tracking mechanism organizes workflow management by explicitly mapping which tasks must complete before others can begin.

Session continuity and task coordination are needed for complex projects requiring sequencing, intent preservation, and clear delegation. When working on features involving research, implementation, review, and testing, tasks ensure progress survives context compaction.

Explicit agent assignment in task subjects works for parallel subagent workflows by using a clear prefix convention. This prefix system ensures clear responsibility when delegating tasks, allowing multiple subagents to operate simultaneously without confusion.

Tasks as persistent working memory require explicit agent assignment in task subjects and the use of dependency tracking for workflow management. This approach applies specifically to complex projects requiring sequencing and is not suited for simple, single-step tasks.