/meeting-prep

name: meeting-prep
description: "Prepare for a specific upcoming meeting by gathering attendee context, history and related topics. Use when the user says 'prep me for my meeting with X', 'what do I need for the 2pm', or before a calendar event. Also use proactively when a meeting is imminent. Not for writing up a meeting that already happened; use `process-meetings`."

Execution mode

Run inline in the current conversation by default, so this work can see what the user has already discussed, decided, or settled this session. Do not fork merely because this skill was selected. Only run in the background when the user explicitly asks for a background run or the host has already obtained a specific background-work approval for this run.

Delegated gathering (large-vault scaling)

This skill stays inline as described above: it keeps session awareness, it asks the user the questions, and it owns every interactive step. What it does NOT do inline is the bulk read-gathering, which on a mature vault (hundreds of notes, thousands of indexed messages, a live calendar and multiple integrations) can be large enough to exhaust the main conversation before the useful work starts.

So the gathering phase is delegated to one `general-purpose` subagent via the Agent tool, using the self-contained prompt in this skill's `AGENT_INSTRUCTIONS.md`:

1. Read `.claude/skills/meeting-prep/AGENT_INSTRUCTIONS.md`. 2. Substitute its placeholders (`{{TARGET_DATE}}`, `{{MEETING_TITLE}}`, `{{ATTENDEE_RECORDS}}`). `{{ATTENDEE_RECORDS}}` is the filtered JSON array from the selected invite, preserving `name`, `person_page`, `email`, `status`, `type`, and `is_current_user` for each attendee. For attendees the user supplied manually, use the same fields with unknown values set to `null`, `type` set to `Person`, and `is_current_user` set to `false`. 3. Call the Agent tool with `subagent_type: "general-purpose"`, that prompt, and a short description. 4. Present its findings as the prep brief, in the Output Format below.

Spot-check before presenting: the brief names person pages, projects and past meetings. Confirm a sample of those files actually exist before repeating their content to the user, and drop anything you cannot stand behind. A subagent's report is a claim, not evidence.

The subagent inherits MCP connections, runs in its own context, and that context is freed when it completes, so only its findings reach this conversation.

**Use `AGENT_INSTRUCTIONS.md` verbatim.** Read the file and pass its content as the subagent prompt, substituting only the placeholders. Do NOT hand-write a replacement brief from what you already know about the meeting: that is how steps get silently dropped, and the omission looks complete because nothing errors. If context from this conversation is worth adding, APPEND it to the file's content; never substitute for it.

**Two caveats that are load-bearing:**

• **Do not count on hooks for the subagent's writes.** The hooks declared in

this skill's own frontmatter belong to this skill's run, not the subagent's, and whether the repository-wide hooks in `.claude/settings.json` reach a subagent's tool calls is not something a skill should assume either way. Nothing in this skill's gathering depends on a hook; the subagent's writes must stand on their own.

• **Always fall back.** If the subagent fails, times out, or returns nothing

usable, say so plainly and run the gathering inline from the same `AGENT_INSTRUCTIONS.md`. A missing subagent must never mean a missing result.

**Stays inline:** confirming which meeting is meant, and presenting the brief.

Prepare for an upcoming meeting by gathering context on attendees and related topics.

Tone Calibration

Before executing this command, read `System/user-profile.yaml` → `communication` section and adapt:

**Career Level Adaptations:**

• **Junior:** Provide more context about attendees, suggest preparation tips
• **Mid:** Balance context with action, suggest talking points
• **Senior/Leadership:** Strategic framing, influence opportunities, key decisions
• **C-Suite:** High-level strategic context, organizational implications, key stakeholders

**Directness:**

• **Very direct:** Bullet points, key facts only
• **Balanced:** Context + talking points (default)
• **Supportive:** Detailed prep, conversation strategies

**Detail Level:**

• **Concise:** Names, roles, top 3 talking points
• **Balanced:** Standard prep format
• **Comprehensive:** Full context, relationship dynamics, strategic considerations

See CLAUDE.md → "Communication Adaptation" for full guidelines.

---

Arguments

**Optional:** $MEETING, $ATTENDEES

If either value is missing, try the calendar before prompting. Ask for the meeting topic or attendee list only when the calendar status or the returned events cannot identify the meeting safely.

**Examples:**

• `/meeting-prep "Q1 Planning" "Sarah Chen, Mike Rodriguez"`
• `/meeting-prep` (then prompt for details)

What This Does

1. Reads the matching calendar invite and keeps its resolved attendee records 2. Looks up only attendees whose invite record has no `person_page` 3. Surfaces recent interactions and open action items 4. Checks for related projects 5. Suggests talking points based on context

Process

Step 0: Gather Context (if needed)

**If $MEETING or $ATTENDEES are not provided, try the calendar before asking.** The user booked the meeting; the invite already holds the title and the attendee list, and asking them to retype it is both friction and a source of duplicate person pages from misspelt names.

When the calendar integration is available, search every calendar visible to Apple Calendar by default:

mcp__calendar-mcp__calendar_get_events_with_attendees(calendar_name="all", start_date="YYYY-MM-DD", end_date="YYYY-MM-DD")

`calendar_name="all"` means all calendars currently visible to the local Apple Calendar integration. It cannot see a calendar ac