SuperAPI AI Directory

long-term-plan

Office & productivity

View repository

Updated 10-16 08:10

About this skill

Long-Term Plan Assistant

Trigger Scenarios

When users need to:

  • Create a multi-day/multi-week long-term project plan
  • Break down a big goal into actionable tasks
  • Schedule tasks to specific dates and set priorities
  • Develop systematic plans in areas such as content creation, study, investment tracking, and more

Workflow

Phase 0: Environment Check and Project Conversion (Execute First)

Important: This is the highest-priority step and must be executed before doing anything else.

Check whether the current conversation is already in a project context:

Case A — Already in a project: If the system prompt contains "project context" information and you can directly call create_tasks、update_plan_document and other tools → skip this phase and go directly to Phase 1.

Case B — Not in a project (regular conversation): If your available tools include convert_to_project, it means this is currently a regular conversation and needs to be converted into a project first:

  1. Directly call the convert_to_project tool, passing in:
    • title: Generate a concise project title based on the conversation content
    • plan_document(optional): Organize the goals and plans already clarified in the conversation into Markdown
  2. After successful conversion,create_tasks、update_plan_document and other project toolsare immediately available, with no need to wait for the next conversation turn
  3. Continue to Phase 1

Note: Do not search for or try other tools to complete the conversion.convert_to_project is the only correct way.

Phase 1: Clarify Goals

Guide the user to answer the following questions:

  • What goal do you want to achieve?
  • What is the time frame? (One week/one month/one quarter)
  • What is your current level or starting point?
  • What known constraints are there? (Time, resources, skills)

Output: A clear project title and core goal description.

Phase 2: Break Down Milestones

Based on the goal and time frame, suggest 3-5 key milestones:

  • Each milestone has a clear, measurable output
  • There are reasonable time intervals between milestones
  • Consider dependencies and learning curves

Use the update_plan_document tool to create a plan document, including:

  • Core goal
  • Milestone list
  • Timeline overview

Phase 3: Decompose Tasks

Break down each milestone into specific tasks, using the create_tasks tool to create them in batches:

  • Each task title is clear and actionable
  • Set a reasonable scheduled_date
  • Distinguish between auto(AI can complete automatically) and manual(requires user participation)
  • Set estimated_minutes estimated time required
  • Establish task dependencies through depends_on_titles Establish task dependencies
  • Set priorityby importance (0=normal, 1=important, 2=urgent)

Phase 4: Confirmation and launch

Show the user the complete plan overview:

  • Total number of tasks, automatic/manual ratio
  • Time distribution
  • Critical dependency path

After confirmation, automatic tasks will begin execution according to the schedule.

Task classification guide

Tasks suitable for auto (automatic execution)

  • Information research and competitive analysis
  • Document drafting and content generation
  • Data collection and organization
  • Template and framework setup
  • Report generation

Tasks suitable for manual (manual execution)

  • Steps that require user decisions
  • Involves operations on external platforms (publishing, registration, etc.)
  • Review and revisions that require subjective judgment
  • Communication and coordination with third parties

Notes

  • Task descriptions should be detailed enough to give the AI clear direction during execution
  • Set dependencies reasonably to avoid circular dependencies
  • Estimated time should take into account the AI's actual execution capability
  • After the initial planning, the plan can be adjusted at any time through conversation