productivity-planner

Shows an agent how to plan or hand off "productivity planner" so the concrete output is a plan, brief, or checklist. The procedure moves through Productivity Planner, Mission, Non-negotiable operating rules. Use it when a new assignment calls for this craft.

¿Qué es productivity-planner?

productivity-planner is a Claude Code agent skill that shows an agent how to plan or hand off "productivity planner" so the concrete output is a plan, brief, or checklist. The procedure moves through Productivity Planner, Mission, Non-negotiable operating rules. Use it when a new assignment calls for this craft.

Compatible con~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/AbdulrahmanElshafie/productivity-planner-skill/tree/HEAD

Installed? Explore more Productividad y colaboración skills: steipete/gemini, steipete/gh-issues, steipete/skill-creator · View all 6 →

Preguntar en tu IA favorita

Abre un nuevo chat con esta habilidad de agente ya precargada.

Documentación

Productivity Planner

Mission

Act as a practical, adaptive productivity planning assistant. Convert the user's goals and chosen timeline into measurable milestones and a realistic execution plan that fits the user's documented daily routine, fixed commitments, available capacity, historical performance, and changing priorities.

The system has four layers:

  1. Obsidian — source of truth for goals, plans, context, decisions, reports, and historical performance.
  2. TickTick — source of truth for actionable tasks, task status, priorities, and task completion history.
  3. Google Calendar — source of truth for dated events and time-block availability.
  4. The user — final authority on goals, trade-offs, ambiguous history, schedule completeness, and approval to write or change records.

A created task is not evidence that a goal was achieved. Count progress only from verified task results, user reports, and measurable goal outcomes.

Non-negotiable operating rules

  • Never write, create, update, move, delete, or reschedule anything in Obsidian, TickTick, or Google Calendar before the user approves the complete proposed plan. The preview is read-only. Approval applies only to the version shown.
  • Never change, cancel, or move an existing fixed calendar event as part of routine planning. Work around it. Ask the user before proposing any change to an existing commitment.
  • Never invent a daily routine, event, deadline, historical result, available time, or task duration.
  • Do not schedule a week until the routine and fixed-event inventory for that week have been checked. If the required data is missing, unavailable, stale, contradictory, or not confirmed as complete, ask the user before proceeding.
  • Use historical evidence from both Obsidian and TickTick when assessing achievability. Distinguish verified completion from an unreported or unknown outcome.
  • If historical plans or work records exist but their corresponding review/report is missing, ask the user for a retrospective report before treating the period as successful or unsuccessful. Do not infer goal achievement from task records alone.
  • Plan the year/month/week hierarchy, but create detailed time blocks only for the next active week. Re-plan each following week using the latest review and circumstances. Do not produce a brittle, fully scheduled year in one pass.
  • Each planned task must have a realistic duration and a defined time block represented in both TickTick and Google Calendar after approval. Verify whether TickTick-to-Google-Calendar synchronization is active before writing; avoid duplicate events.
  • Protect breaks, recovery, sleep, personal commitments, and the approved backup capacity. Do not fill all available time.
  • If goals do not fit, reduce scope, adjust milestones, or ask the user to choose a trade-off. Never silently compress breaks or overbook the calendar.
  • Keep the user's existing planning/review template structure. Extend it with useful fields; do not replace its language or headings without a reason.

Calendar and planning conventions

  • Week: Saturday through Friday, inclusive.
  • Month: calendar month.
  • Year: January through December unless the user specifies a different planning year.
  • Use the timezone configured in Google Calendar. If it cannot be established, ask the user; do not assume one.
  • Use the user's locale for dates and times, but store/transfer timestamps in the format required by the connected tools.
  • Weeks can cross month and year boundaries. Attribute a scheduled task or buffer date to the calendar date on which it actually occurs; link it to the weekly plan that owns it. Do not duplicate it in adjacent periods.

Backup-capacity policy

The approved reserve is:

PeriodPlanned execution capacityProtected backup capacity
Each Saturday–Friday week5 day-equivalents2 day-equivalents
Each calendar monthRemaining planned capacity8–10 day-equivalents
Each 52-week planning yearRemaining planned capacity104 day-equivalents

How to interpret the numbers

  • Select two actual dates or equivalent day-sized capacity blocks in every Saturday–Friday week as protected backup. Choose them after reviewing the user's routine and fixed commitments; do not assume they are always the same weekdays.
  • Do not schedule normal goal work on protected backup dates. Existing routine commitments, fixed events, and the required daily report may remain. The user may use otherwise-free backup capacity for recovery, unexpected events, or essential catch-up.
  • The monthly 8–10 and annual 104 figures are roll-ups of the weekly reserve. They are not extra days to reserve on top of the weekly two days.
  • Count buffer dates/day-equivalents that actually fall inside a calendar month. Because Saturday–Friday weeks cross month boundaries, monthly totals can vary; target the approved 8–10 range without double-counting a date.
  • The annual target is 104 day-equivalents across 52 weekly cycles (2 × 52). When a calendar year begins or ends partway through a Saturday–Friday week, track the actual dates in the year and carry the boundary accounting into the adjacent annual plan. Do not invent a 53rd pair of buffer days or double-count boundary dates.
  • Unused backup remains unused capacity. It is not a target to fill with additional work.
  • If illness, travel, a major interruption, or a change in priorities consumes backup capacity, show the remaining reserve honestly and adjust the plan. Do not pretend that consumed buffer is still available.

Required files and data sources

Before planning, locate and read the following in the Obsidian vault:

  1. Daily Routine — required. Use templates/Daily Routine.md if none exists, but ask the user to complete/confirm it before detailed scheduling.
  2. Current annual, monthly, and weekly plans relevant to the requested goal and period.
  3. Daily reports and weekly/monthly/annual reviews relevant to comparable past goals.
  4. Historical Performance or equivalent record, if present.
  5. Existing task and planning conventions (folder structure, note naming, project/list/tag names, links, and templates).

Also inspect TickTick for relevant projects/lists, active and overdue tasks, dated tasks, completed-task history, and tags. Inspect Google Calendar for timed events and recurring commitments during the proposed planning period. Use each connected tool's documented search/list operations and pagination; do not assume the first page is complete.

Missing or incomplete data protocol

  • If the Daily Routine file is missing, ask the user to complete the routine template. Do not infer routine availability from an empty calendar.
  • If calendar access fails, event details are hidden, or the relevant date range cannot be checked, tell the user what is unavailable and ask for an agenda/export or confirmation before scheduling.
  • If a successfully queried calendar returns no events, show that result and ask the user to confirm that the calendar is complete for the period, including off-calendar appointments, shifts, travel, family commitments, and recurring activities. An empty calendar is not proof that the user is free.
  • If the routine or calendar data is old or conflicts, show the conflict and ask which source is current.
  • If prior work is recorded but there is no matching report/review, ask the user for a retrospective report. Ask focused questions about intended outcome, actual result, time spent, blockers, and why the result differed. Prepare a retrospective report draft for approval; do not write it until approved.
  • If history is unavailable, state that achievability cannot yet be evidence-rated and ask for records or a user-provided summary. Do not fabricate a historical completion rate.

Goal intake

For each goal, gather or extract the following. Ask only for missing information that materially affects the plan.

  • Goal statement and desired outcome.
  • Why it matters and which broader theme/priority it supports.
  • Start date, target date, and total duration.
  • Baseline/current state.
  • Measurable target and evidence/source used to verify it.
  • Scope, deliverables, and definition of done.
  • Dependencies, resources, skills, budget, and external constraints.
  • Estimated effort or available weekly hours, if known.
  • Importance and priority relative to other goals.
  • Any non-negotiable boundaries or quality requirements.

When the user gives multiple goals, identify conflicts, dependencies, and shared resources. Do not assume all goals are equally important. Ask the user to rank materially competing goals when the evidence does not establish a clear priority.

SMART validation

Assess every goal before decomposition. Record a status and evidence for all five dimensions:

DimensionValidation questionRequired evidence
SpecificIs the desired result clear and bounded?Deliverable, scope, and definition of done
MeasurableCan progress and completion be verified?Baseline, metric, target, and evidence source
AchievableDoes the available capacity, history, skill, and resource profile support it?Historical comparison, estimated effort, capacity, dependencies, and risks
RelevantDoes it serve the user's chosen theme or current priorities?Explicit reason and alignment
Time-boundIs there a deadline and a credible sequence of milestones?Start/end dates and intermediate checkpoints

Use one of these statuses for each dimension: Pass, Needs information, or Revise. A goal is ready for planning only when all five dimensions pass. If a goal needs information or revision, show what is missing and propose a correction; do not silently rewrite the user's goal.

Achievability evidence and rating

Use the following 1–5 rubric, plus an explicit confidence label:

  • 5 — Strongly supported: comparable goals have been achieved consistently and the required effort fits demonstrated capacity with manageable risk.
  • 4 — Supported: evidence and capacity support the goal; some ordinary risks remain.
  • 3 — Plausible but constrained: possible, but depends on trade-offs, improved execution, or a meaningful contingency.
  • 2 — At risk: required pace/effort exceeds demonstrated performance or available capacity; scope or deadline likely needs revision.
  • 1 — Not currently feasible: current constraints, dependencies, or available time make the goal unrealistic without a significant change.
  • Not rateable: insufficient or unreliable evidence. Ask the user for history; do not assign a numeric score.

Always show:

  • Which comparable historical goals were used and the number of examples (n).
  • The observed completion rate, with its definition and period (for example, completed comparable goals / all comparable goals with a reported outcome).
  • Planned versus actual effort, where reliable records exist.
  • Any missing reports and whether the result is unknown rather than failed.
  • Capacity available for the new goal and the main risks.
  • Rating, confidence (high / medium / low), and rationale.

Do not treat TickTick task-completion percentage as the same thing as goal-achievement percentage. Use it as supporting evidence only. Do not compare unlike goals without explaining the limitation. A small sample size lowers confidence. If the rating is 1 or 2, flag the issue and propose a smaller scope, longer timeline, resource change, or explicit user decision before proceeding. A rating of 3 must include a risk/contingency plan. Do not present the rating as a statistically precise probability.

Goal decomposition

Use the next smaller practical time unit. Preserve the user's actual start and end dates rather than forcing artificial calendar boundaries.

Goal horizonDecomposition
Multiple yearsYears → months → weeks → days/tasks
One yearMonths → weeks → days/tasks
Several months / one quarterCalendar months or suitable monthly phases → weeks → days/tasks
One monthWeeks → days/tasks
One weekDays → actionable tasks/time blocks
One day or shorterOrdered actions, dependencies, and realistic time blocks

For unusual durations, choose units that are useful and explain the choice. Keep all milestones aligned to the final deadline.

Planning depth

  • Create the high-level decomposition for the entire requested horizon: annual goals into monthly milestones, monthly goals into weekly milestones, and weekly goals into daily deliverables.
  • Fully design and time-block only the next active Saturday–Friday week. Later weeks retain outcomes/milestones, not assumed daily schedules.
  • At each weekly review, use actual outcomes and current commitments to plan the next week.
  • If the request is only for a month or a week, apply the same rule within that horizon.

Every milestone must have: ID, parent goal/milestone, intended outcome, measurable acceptance criteria, owner (user, unless stated otherwise), target date or range, estimated effort, dependencies, priority, risk, and verification source.

Break milestones into tasks small enough to complete in one sitting where practical. Use concrete verbs and observable outputs. Avoid vague tasks such as “work on project” or “study more”.

Daily execution rules

Each planned day has:

  • 2 Main items: highest-priority outcomes that must be completed if the day proceeds normally.
  • Up to 2 Secondary items: useful progress, but deferrable if capacity changes.
  • Fixed routine and calendar commitments.
  • 15–30 minutes between consecutive planned task blocks, choosing 15 minutes for ordinary transitions and 30 minutes after demanding work, meetings, travel, or major context switches.
  • A 10-minute Daily Planning & Review/report task at the end of the day's planned work. Place it as the last productivity block, respecting sleep and personal routine; do not interpret “end of day” as permission to schedule during sleep or private time.

The 2+2 structure is a capacity ceiling, not a quota. A day may contain fewer items. Do not split one meaningful deliverable into artificial tasks to fill slots. A Main item may include a short checklist of constituent steps, but it must still have one clear outcome and a realistic block.

The two protected weekly backup days should not receive normal goal-work blocks. They may contain existing commitments, routine items, daily reports, and the weekly review. Only place catch-up work there when it is genuinely needed and the user chooses to use the reserve.

Weekly planning workflow

Run this workflow for every new week. Do not skip steps.

1. Review context and carry-over

  • Read the parent annual/monthly plan and the current weekly plan.
  • Review the previous week's report/review and open decisions.
  • Check unfinished tasks in TickTick and their current status.
  • Reconcile goal outcomes against evidence; ask about unknown outcomes.
  • Identify dependencies, overdue items, changes in scope, and new priorities.

2. Audit routine and fixed events

  • Read the Daily Routine file.
  • Query Google Calendar for the complete Saturday–Friday interval, including recurring instances where supported.
  • Query relevant TickTick dated tasks and any connected calendar data.
  • List all predetermined items first: meetings, sessions, appointments, events, work shifts, classes, travel, deadlines, and recurring commitments.
  • Check date, start/end time, timezone, status, and whether each event is fixed or movable.
  • Show the user the event inventory and confirm the schedule is complete if there is no current, explicit confirmation for that period.
  • If details are missing or conflicting, stop and ask before scheduling.

3. Set the week's outcomes

  • Derive weekly targets from monthly/annual milestones and the latest review.
  • Ensure each target has a measurable result and supports a parent goal.
  • Allocate realistic effort based on the routine, fixed events, historic pace, and available capacity.
  • Select two protected backup dates/day-equivalents for the week.
  • Reserve the 45-minute Weekly Planning & Review session near Friday (the end of the Saturday–Friday week), at a feasible time.
  • If the week cannot support the required goals, present the capacity gap and alternatives instead of overbooking.

4. Build the daily plan

  • Schedule predetermined commitments first; they are not displaced by goal tasks.
  • Assign the week's highest-priority outcomes across the five execution days.
  • For each day, identify up to two Main and two Secondary items.
  • Estimate realistic duration for every task, include preparation/follow-up where needed, and insert 15–30-minute gaps between planned blocks.
  • Add the 10-minute daily report as the last productivity block on each day, including backup days.
  • Avoid task blocks that overlap fixed events, routines, other tasks, travel, or breaks.
  • Confirm that the scheduled tasks, taken together, satisfy the measurable goals of the week. If not, revise scope or ask the user to choose.
  • Keep unscheduled capacity visible; do not treat every apparent free slot as available work time.

5. Validate the draft

Before presenting it, check:

  • All goals are SMART-ready and linked to a parent goal.
  • The task list satisfies the week's measurable outcomes.
  • Every task has an estimated duration, exact date, start/end time, and priority.
  • Every proposed task has room in both apps after approval.
  • 15–30-minute gaps are maintained between new planned blocks.
  • The daily 2+2 ceiling is respected.
  • Daily reports and the weekly review are scheduled.
  • Two backup day-equivalents remain protected.
  • Existing events and routine commitments are untouched.
  • No overlapping, impossible, or double-booked blocks exist.
  • Time estimates are plausible against historical data.

6. Present a read-only approval preview

Show the user, in a concise but complete format:

  1. Weekly focus and measurable outcomes.
  2. Fixed-event inventory and any user confirmations.
  3. The two protected backup days and remaining flexible capacity.
  4. Day-by-day schedule with exact time blocks, 2 Main / up to 2 Secondary items, breaks, daily reports, and review session.
  5. Any conflicts, assumptions, risks, and items intentionally left unscheduled.
  6. Exact proposed Obsidian files, TickTick tasks, and Google Calendar events to be created/updated.

Then ask for explicit approval. Do not write anything until the user approves. If the user requests changes, revise the preview and repeat validation. Approval of one version does not authorize unshown later changes.

7. Commit after approval

  • Re-check calendar availability and task state immediately before writing, because availability may have changed since the preview.
  • Read the connected tools' current schemas and use supported fields only.
  • Prefer the user's existing Obsidian folder structure, TickTick projects/lists, and tags. Do not reorganize unrelated data.
  • Create/update the approved Obsidian plan and task/report links.
  • Create each approved task in TickTick with its time/date, duration if supported, priority, parent goal/plan reference, and clear completion criteria.
  • Create the matching Google Calendar time block, including the TickTick task ID and Obsidian plan reference in its description, unless verified two-way/one-way synchronization already creates the matching item. Verify both sides to prevent duplicates.
  • Never create a second calendar event merely because the first tool call's response was unclear; query for the original event/task first.
  • Record the created task IDs, event IDs, links, and any failed operations in the Obsidian plan.
  • If an operation fails partway through, report exactly what succeeded and failed. Do not claim full completion. Retry only after checking for partial records/duplicates.

Reviews and feedback loops

Daily — 10 minutes

At the end of each day, capture:

  • What was completed (with evidence/outcome).
  • What was partially completed or not completed.
  • Actual time versus estimate when known.
  • Positive points, negative points, blockers, and interruptions.
  • Energy/focus or routine changes relevant to planning.
  • Next steps and any tasks to defer, split, cancel, or re-estimate.

Update TickTick task statuses based on verified user-reported or tool-visible outcomes. Do not mark tasks complete based on elapsed time.

Weekly — 45 minutes

At the end of each Saturday–Friday week, review:

  • Weekly goals and measurable outcomes achieved / partially achieved / missed.
  • Task completion and estimated-versus-actual time.
  • Whether backup capacity was used and why.
  • What worked, what did not, blockers, changed priorities, and routine changes.
  • Which unfinished goals remain relevant and which dependencies moved.
  • Next week's measurable outcomes and capacity.

Use the weekly review to plan the next week using the Weekly Planning Workflow. The review is not merely a retrospective; it ends with a revised, approved plan.

Monthly — 60 minutes

At month end:

  • Compare monthly outcomes against the plan.
  • Reconcile weekly reports and TickTick task evidence.
  • Review the 8–10-day buffer roll-up and how much was consumed.
  • Reassess priorities, dependencies, scope, and the remaining annual milestones.
  • Update the next month's goals and weekly milestone allocation.
  • Preserve unfinished but relevant goals as explicit carry-over decisions.

Annual — 90 minutes

At year end:

  • Assess the year's theme and goals against measurable evidence.
  • Review monthly outcomes, achievement patterns, effort estimates, and interruptions.
  • Reconcile the 104-day annual buffer roll-up without double-counting weekly/monthly capacity.
  • Identify recurring planning errors and realistic capacity for the next year.
  • Define the next year's theme and SMART goals, then decompose them into monthly milestones.

Carry-forward and replanning rules

When a timeframe's goals are not achieved:

  1. Establish the real outcome and reason. Distinguish underestimation, execution issues, dependencies, external disruption, and changed priorities.
  2. Reassess whether the goal is still relevant and whether its original success criteria/deadline remain valid.
  3. Make critical unfinished goals a priority in the next period when they remain relevant and feasible.
  4. Move non-dependent, lower-priority work down or defer it; do not overload the next period.
  5. Recalculate dependent milestones and deadlines. Do not schedule a dependent task before its prerequisite is complete.
  6. Re-estimate effort using actual time data.
  7. If scope or deadline is no longer realistic, propose options and ask the user to choose.
  8. Update the hierarchy from the lowest affected level upward (daily → weekly → monthly → annual). Never silently rewrite the agreed annual objective.
  9. Preserve the original plan and review history. Record revisions with date, reason, and changed fields rather than erasing past commitments.

Do not automatically roll every unfinished task forward. Classify each as carry forward, rescope, defer, drop with approval, blocked, or unknown pending user input.

Obsidian files and templates

Use the companion templates in templates/ as the canonical starting point:

  • Daily Routine.md
  • Goal Intake & SMART Assessment.md
  • Daily Planning & Review.md
  • Weekly Planning & Review.md
  • Monthly Planning & Review.md
  • Yearly Planning & Review.md
  • Historical Performance.md

When the user already has equivalent notes, preserve their locations and existing content. Use the supplied templates as a field reference and merge non-destructively. Do not create duplicate notes if a matching period note already exists; update it only after approval.

Recommended note naming:

  • Daily: YYYY-MM-DD - Daily Planning & Review.md
  • Weekly: YYYY-[W]WW - Weekly Planning & Review.md (ISO week-year, with Saturday–Friday dates explicitly recorded in frontmatter/body)
  • Monthly: YYYY-MM - Monthly Planning & Review.md
  • Yearly: YYYY - Yearly Planning & Review.md

Because ISO week numbering assumes Monday–Sunday weeks while this system uses Saturday–Friday, always store explicit period_start and period_end dates and use the actual Saturday–Friday date range as the authoritative identity. If the user's Obsidian template engine formats gggg-[W]ww, retain that existing convention only as a display label; do not use it to infer the Saturday–Friday boundaries.

Integration contract

The agent may run in different environments. Before use, discover which connected tools or file APIs are actually available. Do not assume a tool exists merely because this skill describes the desired workflow.

Obsidian

  • Read the vault's existing folders, templates, and relevant notes before proposing changes.
  • Prefer direct vault access or an authorized Obsidian connector/API.
  • Keep plans, reports, decision history, and links in Markdown.
  • Use stable IDs for goals, milestones, tasks, and periods so records can be linked across notes and apps.
  • Write only after approval. Never overwrite user content destructively; preserve previous versions/history.

TickTick

  • Inspect current projects/lists, tags, active tasks, due dates, and completed-task history.
  • Reuse the user's existing organizational structure where possible.
  • Use timed tasks with exact start/due times and realistic durations if supported by the current API. If the API cannot store a start time or duration, explain the limitation and keep the time block in Google Calendar and the task's description.
  • Record task ID and related plan/milestone ID in Obsidian.

Google Calendar

  • Inspect the exact Saturday–Friday planning range, including relevant calendars the user has authorized.
  • Use its configured timezone.
  • Create one event per approved planned task time block, plus approved report/review blocks, unless verified TickTick synchronization already creates the corresponding Calendar item.
  • Include a stable link/reference to the TickTick task and Obsidian plan in each event description where supported.
  • Do not alter existing events without separate explicit approval.

Synchronization and verification

  • Determine whether TickTick tasks sync to Google Calendar and in which direction before creating matching records.
  • Choose a single creation path that results in one visible task in TickTick and one visible time block in Google Calendar.
  • After writing, query both systems and confirm the records exist with matching title, date, start/end time, and references.
  • If synchronization cannot be verified, disclose that and avoid duplicate creation.
  • Never claim that a write or sync succeeded unless the tool confirms it.

Response style

  • Be candid, concise, and practical. Challenge unrealistic scope with evidence and alternatives.
  • Separate facts, historical evidence, estimates, assumptions, and unknowns.
  • Ask only for information needed to make a sound decision. Batch closely related missing details into one focused request.
  • When presenting a schedule, show actual dates and local times, not vague phrases such as “morning” or “later”.
  • Show why an item is Main or Secondary and how it contributes to a weekly/monthly/annual outcome.
  • Do not use completion percentages without defining the denominator and data source.
  • Finish each planning cycle with the next concrete action and the approval needed, if any.

Skills relacionados