Timesheet Builder

Use ISO week numbering for internal tracking, but display dates in human-readable format (e.g., "Week of 24-28 Mar 2026"). Weekend work hours are attributed to that week's totals.

Step 2: Find or create the timesheet file

Look for an existing timesheet markdown file in the user's workspace folder. Check for files named timesheet.md, timesheets.md, or similar. If found, read it to extract:

• Categories/allocations: The list of project names and internal categories already in use
• Client mappings: Infer from previous weeks which email domains and meeting patterns map to which categories (e.g., @airbnb.com attendees = an Airbnb project, @kcl.ac.uk = Kindred)
• Week numbering: Continue from where the file left off
• Formatting conventions: Match the existing table structure and notes style

If no file exists, tell the user you'll create one after gathering enough data to establish categories. Don't create the file until after the first week's data has been categorised and confirmed.

Step 3: Pull data from all sources

For each week, pull data from all available sources in parallel. Not all sources may be available; use whatever the user has connected.

Google Calendar

Use gcal_list_events with:

• timeZone: Ask the user their timezone on first run, or infer from calendar settings
• condenseEventDetails: true to keep responses manageable
• maxResults: 100 (increase if the user has very busy weeks)
• Time range: Monday 00:00 to Sunday 23:59 of the target week (include the following weekend, as work sometimes spills into Saturday/Sunday and should be attributed to that week)

If the response is too large and gets persisted to a JSON file, parse it with a Python script. The persisted format is [{"type": "text", "text": "<JSON string>"}] at the top level.

For each event, extract:

• Summary (meeting title)
• Start/end times (for duration calculation)
• Attendees and their email domains
• Organiser
• Status (ignore cancelled events)
• Event type (skip workingLocation and focusTime events)

Slack

Use slack_search_public_and_private with:

• from:<@USER_ID> to find the user's messages
• Date filters for the target week
• sort: timestamp, sort_dir: asc
• include_context: true for surrounding conversation

Slack messages help fill gaps: they reveal what someone was working on between meetings, show async collaboration, and surface things like "I worked late on X" or overtime indicators. They also help disambiguate meetings (e.g., a 1:1 where the Slack context reveals it was about a specific project).

Exclude personal/social messages from time allocation.

Granola meeting notes

Use list_meetings with a custom date range covering the target week (Monday through Sunday). This provides meeting titles, participants with email addresses, and dates. Cross-reference with calendar events to validate attendee lists.

For meetings that are ambiguous or particularly important, use get_meetings to pull the full meeting notes. Granola notes often contain rich context about what was discussed, decisions made, and action items assigned, all of which help categorise the meeting more accurately than the title alone. Prioritise pulling detailed notes for: meetings with vague titles (e.g., "Quick sync", "Catch up"), meetings where the attendee list doesn't clearly indicate a project, and meetings that appear in multiple potential categories.

Gmail (if available)

Use gmail_search_messages to search for work-related email threads during the period. Email can reveal work that doesn't show up in calendar or Slack, like vendor communications, client correspondence, or document reviews.

Step 4: Categorise time

This is the most important and nuanced step. The goal is to assign every working hour to a category.

How to categorise meetings

Use these signals in priority order:

1. Attendee email domains: External domains strongly indicate client work. Build a domain-to-client mapping from the existing timesheet or by asking the user. Internal company domains (e.g., @raredays.com, @coursestudio.com, or whatever the user's org uses) indicate internal time. When you see a mix of internal and external attendees, the external domain usually determines the client project. For example:

   • @clientcompany.com attendees = that client's project
   • All-internal attendees = internal category

2. Meeting title patterns: Many recurring meetings have clear category signals:

   • "Standup", "Sprint", "Retro", "Refinement" with a project name = that project
   • "L10", "Leadership", "All Hands", "Resourcing" = Internal - Leadership
   • "Interview", "Hiring", "Dev Sync", "Eng Meeting" = Internal - Engineering
   • "1:1", "/ " separator between two names = Internal - 1:1 (if both internal)
   • Titles prefixed with "RDA:" or similar company-wide prefixes = Internal (the specific internal category depends on the meeting content, e.g., "RDA: Leadership L10" = Leadership, "RDA: Engineering L10" = Engineering, "RDA: All Hands" = Leadership)

3. Slack channel context: Messages in #int-projectname or #ext-projectname channels help attribute async work.

4. Granola participant lists: Confirm who was actually in a meeting (calendar invites don't always reflect attendance).

Duration rounding

Round meeting durations up to the nearest half hour. A 25-minute meeting = 0.5hrs. A 35-minute meeting = 0.5hrs. A 50-minute meeting = 1hr. A 65-minute meeting = 1.5hrs.

Handling overlapping meetings

Calendar frequently shows overlapping events. The user can only attend one at a time. When you spot overlaps, ask the user which they attended, or if they split time between them.

Filling unscheduled time

Calendar meetings rarely account for a full 8-hour day. The gap between meeting hours and the daily target represents unstructured work time: deep work, code reviews, document writing, Slack discussions, email, etc.

Distribute unstructured time based on:

• Slack activity during those hours (what channels were they active in?)
• The overall weight of each project that week (if 40% of meetings were for Project X, lean toward allocating more gap time there)
• Ask the user if the distribution looks right

1:1 meetings

Two-person meetings where both attendees are from the user's company default to "Internal - 1:1". However, flag any 1:1 where Slack context suggests it was primarily about a specific project, and ask the user if it should be recategorised.

Special day types

• Stat holidays: Look for all-day "STAT HOLIDAY" or "Out of office" events on known holidays. Allocate 8hrs to "Stat Holiday".
• PTO/vacation: Look for all-day OOO events. Allocate to "PTO".
• Partial days: If someone was OOO for part of a day, reduce that day's hours accordingly.

Step 5: Ask clarifying questions

Batch your questions efficiently using the AskUserQuestion tool (up to 4 questions per batch). Typical things to ask:

• Unknown attendees: "Meeting X has attendees from @newdomain.com. What client/project is this?"
• Ambiguous meetings: "Quick AI huddle with Tonner, Miles, Bruno. Is this Leadership or Engineering?"
• Overlapping events: "KTB Sync and Airbnb AI Chat overlap at 1:30pm Wednesday. Which did you attend?"
• 1:1 exceptions: "Your 1:1 with Bhimesh on Wednesday seems to be about the Airbnb project. Should this be Airbnb Supply instead of 1:1?"
• Overtime: "Your meetings total 32hrs but your target is 40. Any overtime this week, or should I distribute the remaining 8hrs across your active projects?"
• Personal events to exclude: "I see Dentist on Monday and Sunny Run on Wednesday. Confirm these should be excluded?"

On the first run with no existing mappings, also confirm the full category list with the user before writing anything.

Step 6: Confirm weekly hours

Ask the user to confirm:

• Target hours per week (default 40)
• Whether there was any overtime this week, and if so how many extra hours and where to allocate them
• Whether any weekend work was done (Saturday/Sunday data is pulled by default; if work is found, confirm with the user how to allocate it and whether it counts as overtime)

Step 7: Write the timesheet

File structure

If creating a new file:

# [User's Name]'s Timesheet - [Company]

Started: [First date]

## Categories

- [Category 1]
- [Category 2] (with any clarifying notes)
- ...

## Mappings

These mappings help categorise future weeks automatically:

- @clientdomain.com = Client Project Name
- "Meeting Title Pattern" = Category
- channel:#slack-channel = Category
- person:[email protected] = Category (for people strongly associated with one project)

---

## Week N (DD-DD Mon YYYY)

| | Mon DD | Tue DD | Wed DD | Thu DD | Fri DD | Total |
|---|---|---|---|---|---|---|
| Category 1 | X | X | X | X | X | XX |
| Category 2 | X | X | X | X | X | XX |
| **Total** | **8** | **8** | **8** | **8** | **8** | **40** |

**Notes:**
- Key activities, meetings, and context for the week
- What was discussed, who was involved, any notable events
- Excluded personal items

Mappings section

The Mappings section is important because it persists learned associations between runs. Each time a new client domain, meeting pattern, or person-to-project association is confirmed by the user, add it to this section. On subsequent runs, read these mappings first to reduce the number of questions needed.

Table formatting rules

• Categories as rows, weekdays as columns, with a Total column
• Only include categories that have hours allocated (don't show zero rows)
• Bold the Total row
• Hours in half-hour increments (0.5, 1, 1.5, etc.)
• Each day should sum to the target (usually 8)
• If overtime, the weekly total will exceed 40

Notes section

Write concise but informative notes that explain what the numbers represent. Include:

• Key meetings and what they covered
• Notable Slack activity or async work
• Overtime explanation if applicable
• Excluded personal items
• Any data source gaps (e.g., "No Slack data available for Thursday")

Edge cases

• Short weeks: If a week has fewer than 5 working days (holidays, PTO), adjust the table to show only working days, or show all 5 with holiday/PTO rows filling non-working days.
• Very meeting-heavy days: When calendar meetings exceed the daily target (e.g., 9+ hours of meetings in an 8-hour day), check whether the day started earlier or ended later than normal working hours. If meetings span, say, 7am to 6pm, the user likely attended most of them and it was an overtime day rather than a case where meetings were skipped. In that scenario, suggest it as a potential overtime day and ask the user to confirm. Only assume meetings were skipped if the meeting hours exceed the daily target but all fall within normal working hours (suggesting double-booking).
• No data from a source: If one of the data sources isn't connected or returns nothing, note it and rely more heavily on the other sources. Don't fail silently.
• First-time users: If there's no existing timesheet and no established categories, spend extra time on Step 4/5 to build the category list from scratch. Pull a week of data, present your best guess at categories based on the domains and meeting patterns you see, and have the user confirm/adjust before writing anything.

Tips for accuracy

• Cross-reference sources: A calendar event + Slack messages about the same topic = high confidence in categorisation
• Watch for "Block" events in calendar: These are often placeholder deep-work blocks. Check Slack activity during those times to infer what was actually worked on.
• Recurring meetings with the same title may belong to different projects across weeks if team assignments change. Don't assume; verify if something looks off.
• When in doubt, ask. A quick clarifying question is better than a wrong allocation the user has to fix later.