Meeting Prep Workflow

You're helping the user prepare for an upcoming meeting by gathering relevant context from their Obsidian vault.

Context

Today's date is {{CURRENT_DATE}}. The vault is at /workspace.

The user manages 5 direct reports: Armando, Christina, Helio, Jakob, Shaun. The user's boss is Thomas. Peers: Maggie (Magdala, project manager), Que (business analyst).

External Resources:

• •ITSE Playbook: /workspace/playbook/ - Contains SOPs and best practices. SOPs at /workspace/playbook/docs/standard-operating-procedures/, best practices at /workspace/playbook/docs/fundamentals/
• •GitHub Wiki: /workspace/github-wiki/ - Contains developer meeting agendas and notes named Developer-Meeting-Agenda-YYYY-MM-DD.md
• •Jira API: Jira helper functions available at /workspace/.claude/jira-helper.sh for fetching live ticket status

Jira Integration

Source the Jira helper to fetch live ticket status:

source /workspace/.claude/jira-helper.sh

# Extract all ticket codes from the meeting file
tickets=$(jira_extract_tickets "dev meeting.md")

# Fetch status for each ticket
jira_get_multiple_tickets "$tickets"

When to use Jira:

• •When Jira tickets (UP-, RWSP2-, WEBP4-, UCDCSISDEV-, UCDCWEB2-, RWR-) are mentioned in meeting notes or action items
• •Use jira_extract_tickets to find all ticket codes in the meeting file
• •Use jira_get_multiple_tickets to fetch their current status
• •Include Jira status in a dedicated "Referenced Tickets" section
• •Especially useful for planning, refinement, and dev meetings where tickets are frequently discussed

Step 1: Resolve Meeting File

The user provides freeform input (e.g., "Christina", "dev meeting", "portal planning"). Resolve it to a meeting file using this priority order:

1. •Exact match: {input}.md
2. •1-on-1 pattern: {input} 1on1.md
3. •Meeting pattern: {input} meeting.md
4. •Planning patterns: {input} planning.md, {input} planning meeting.md
5. •Glob fallback: *{input}* (case-insensitive)
6. •Ambiguous: If multiple matches found, use AskUserQuestion to let the user choose

Use Glob to check each pattern. Stop at the first pattern that finds exactly one match.

Special case: If the input is a person's name (e.g., "Christina"), prefer the 1-on-1 file (Christina 1on1.md) over other matches.

Step 2: Detect Meeting Type

Once the file is found, classify it by matching the filename against these patterns:

Step 3: Extract Previous Meeting Content

Queued notes: Any content at the top of the file before the first ## dated heading is considered "queued discussion items" — notes the user jotted down between meetings for the upcoming session. Capture this content separately; it will be surfaced in Step 7 and incorporated by Step 8.

Read the meeting file and find the most recent dated section. Look for heading patterns like:

• •## YYYY-MM-DD
• •## Month DD, YYYY
• •## M/D/YYYY or ## MM/DD/YYYY
• •## YYYY-MM (monthly)
• •Any heading containing a recognizable date

Extract the content under the most recent dated heading as the "previous meeting summary." Summarize it briefly (3-5 bullet points covering key decisions, action items, and discussion topics).

Also note the date of that section — this becomes the "last meeting date" used in Step 5.

Step 4: Find Open Tasks

Search the meeting file for incomplete tasks:

• •Pattern: - [ ] (lines starting with - [ ])
• •Exclude completed (- [x]) and cancelled (- [-])

Format each task with its file path and line number reference.

Also check for tasks with due dates (📅 YYYY-MM-DD) and flag any that are overdue relative to today.

Step 5: Gather Cross-Vault Context

Search for mentions of the person or project in other vault files since the last meeting date.

For 1-on-1 meetings (direct-report, boss, peer):

• •Grep for the person's name across the vault (excluding the meeting file itself)
• •Focus on mentions in: dev meeting.md, other 1-on-1 files, project files, and planning files
• •Filter to entries dated after the last meeting date when possible

For dev meetings:

• •Check each direct report's 1-on-1 file for recent entries (since last dev meeting)
• •Look for blockers, completed work, and upcoming items
• •GitHub Wiki agenda: Look for today's agenda at /workspace/github-wiki/Developer-Meeting-Agenda-YYYY-MM-DD.md. If not found, find the most recent one by globbing Developer-Meeting-Agenda-2026-* and sorting. Surface any pre-populated topics or carryover items from the wiki agenda.
• •Playbook SOP: Read /workspace/playbook/docs/standard-operating-procedures/conduct-a-developer-meeting.md and remind the user of the standard flow

For planning meetings:

• •Search for the project prefix (UP-, RWSP2-, WEBP4-, UCDCSISDEV-, UCDCWEB2-, RWR-) across the vault
• •Find recent decisions, blockers, and status updates
• •Playbook: Search /workspace/playbook/docs/ for SOPs relevant to the project or discussion topics (e.g., deploy, release, security)

For general/other meetings:

• •Search for keywords from the meeting filename across the vault
• •Find recent mentions and related tasks
• •Playbook: Search /workspace/playbook/docs/ for SOPs matching the meeting topic keywords

Step 6: Build Suggested Discussion Topics

Synthesize suggestions from what you've gathered:

• •Open tasks that need follow-up or are overdue
• •Recent mentions of the person/project in other contexts
• •Unresolved items from the previous meeting
• •Cross-cutting concerns (items appearing in multiple files)
• •Overdue items that need escalation or rescheduling

Limit to 3-7 concrete, actionable topics.

Step 7: Format and Display Output

Before formatting output: Extract any Jira ticket codes (UP-, RWSP2-, WEBP4-, UCDCSISDEV-, UCDCWEB2-, RWR-) mentioned in the meeting file and fetch their current status using the Jira helper.

Output the prep summary in this structure:

## Meeting Prep: {meeting name}
**Type**: {detected type}
**File**: {filepath}
**Last entry**: {date of most recent section}

### Queued Discussion Items (if any)
{Content found before the first dated heading — notes the user added between meetings for this session. Display verbatim or lightly formatted.}

### Previous Meeting Summary
{Brief summary of last dated section — 3-5 bullet points}

### Open Action Items
{Incomplete tasks from the meeting file, with file:line references}

### Referenced Tickets (if any Jira tickets found)
{Live status of Jira tickets mentioned in the meeting notes}
{Use jira_ticket_summary for each ticket}

### Context from Other Notes
{Mentions of this person/project in other vault files since last meeting}

### Suggested Discussion Topics
{3-7 derived topics from open tasks, recent mentions, overdue items}

Type-Specific Additions

For direct report 1-on-1s (direct-report type), add this section after Suggested Discussion Topics:

### People / Product / Process Prompts
- **People**: How are things going with the team? Any collaboration challenges?
- **Product**: What's the most important thing you worked on since we last met? Any blockers?
- **Process**: Is there anything slowing you down or that we could improve?

For boss 1-on-1 (boss type), add:

### Questions to Bring
{Derived from open items, decisions needed, and things requiring Thomas's input}

For dev meetings (dev-meeting type), add:

### GitHub Wiki Agenda
{Topics and action items from the wiki agenda for today's meeting, or the most recent agenda if none exists for today. Include the wiki filename for reference.}

### Team Status (from recent 1-on-1s)
{For each direct report, show 2-3 bullet points from their most recent 1-on-1 entry}

### Meeting Flow (from Playbook)
{Key steps from the playbook SOP for conducting a developer meeting}

For planning meetings (planning type), add:

### Referenced Tickets
{Extract tickets from the meeting file using jira_extract_tickets}
{Fetch live status using jira_get_multiple_tickets}
{Show ticket code, summary, status, assignee for each}

### Vault References
{Tasks and mentions matching the project prefix from other vault files}

### Risk Items
{Overdue tasks, blocked items, items mentioned as risks}

For IT monthly (it-monthly type), add:

### Team Summary for Report
{Recent accomplishments from dev meeting notes and 1-on-1s that could be shared}

Step 8: Offer to Scaffold Meeting Section

After displaying the prep summary, offer to create a new dated section in the meeting file for today's meeting.

Check for Existing Section

First, check whether the meeting file already has a section for today's date (## {{CURRENT_DATE}}). If it does, skip this step entirely — do not offer to scaffold.

Ask the User

Use AskUserQuestion to ask:

Want me to add a new section for today's meeting?

Options: Yes / No

If the user declines, end the workflow.

Determine Template

If the user accepts, insert a new section based on the meeting type detected in Step 2:

1-on-1 types (direct-report, boss, peer) — use the structure from templates/1on1.md:

## {{CURRENT_DATE}} — {Person's Name}

### People
-

### Product
-

### Process
-

### Action Items
- [ ]

Dev meeting (dev-meeting) — use the structure from templates/dev-meeting.md:

## {{CURRENT_DATE}}

### Standup
- did before
	-
- will do next
	-

### Show & Tell
-

### Discussion
-

### Action Items
- [ ]

All other types (planning, refinement, it-monthly, general) — insert a bare heading:

## {{CURRENT_DATE}}

Handle Queued Notes

The user may have added loose notes (discussion items, reminders, links) to the top of the file before the first ## dated heading, intended for the upcoming meeting. These queued notes must be absorbed into the new section, not left stranded above it.

• •If there is content before the first ## heading, move it into the new section — place it right after the ## date heading line (before the template subsections like ### People).
• •This means the old_string for the Edit should start from the beginning of the queued content and extend through the first ## heading line.

Insert Location

Insert the new section before the first existing ## dated heading. This matches the vault convention of newest-first ordering.

• •Use the Edit tool. The old_string should be the first ## heading line in the file (or, if queued notes exist, start from the first line of queued content through that heading). The new_string should be the new dated section (with any queued notes folded in) followed by that same original ## heading line.
• •If the file is empty or has no ## headings, append the section at the end of whatever content exists.

Confirm

After inserting, confirm to the user what was added and where. If queued notes were absorbed, mention that too, e.g.:

Added a new ## 2026-02-11 — Christina section with People/Product/Process/Action Items to the top of Christina 1on1.md. Your 3 queued discussion items were moved into the new section.

Important Implementation Notes

Scaffolding

• •The only file modification this skill performs is inserting a new dated section via the Edit tool in Step 8
• •Only modify the meeting file identified in Step 1 — never create new files or edit other files
• •Always ask the user before making any changes
• •All other output (the prep summary) is displayed directly to the user

File References

• •Always include filepath:line_number references so the user can navigate to sources
• •Use relative paths from the vault root when displaying

Date Handling

• •Parse dates in YYYY-MM-DD format (most common in this vault)
• •"Since last meeting" means entries dated after the last meeting date found in Step 3
• •If no date can be determined, include all recent content (last ~30 days)

Search Scope

• •Search the entire vault but skip .obsidian/, Omnivore/, Clippings/, and Learning from Dutch/ directories
• •Prioritize files that are likely related (same project, same person, meeting files)

User Experience

• •Be concise — this is a prep summary, not a full report
• •Highlight what's actionable and what needs attention
• •If there are no open tasks or recent mentions, say so briefly rather than padding
• •If the meeting file is empty or very short, note that and focus on cross-vault context

Example Session Flows

Direct Report 1-on-1

User: /meeting-prep Christina

1. Resolve: finds "Christina 1on1.md"
2. Detect: direct-report type
3. Read file, find last dated section
4. Find open tasks in Christina's file
5. Search vault for "Christina" mentions since last meeting
6. Generate prep with People/Product/Process prompts

Dev Meeting

User: /prep dev meeting

1. Resolve: finds "dev meeting.md"
2. Detect: dev-meeting type
3. Read file, find last dated section with action items
4. Find open tasks in dev meeting file
5. Check each direct report's 1-on-1 for recent updates
6. Read today's GitHub wiki agenda (Developer-Meeting-Agenda-YYYY-MM-DD.md)
7. Read playbook SOP (conduct-a-developer-meeting.md) for meeting flow
8. Generate prep with wiki agenda, team status, and meeting flow sections

Planning Meeting

User: /prep portal planning

1. Resolve: finds "portal planning.md"
2. Detect: planning type, project = "portal" (UP- prefix)
3. Read file, find last dated section
4. Find open tasks in planning file
5. Search vault for UP- ticket references
6. Generate prep with open tickets and risk items sections

Boss 1-on-1

User: /prep Thomas

1. Resolve: finds "Thomas 1on1.md"
2. Detect: boss type
3. Read file, find last entry
4. Find open tasks
5. Search for items needing Thomas's input
6. Generate prep with "Questions to Bring" section

References

• •https://justoffbyone.com/posts/how-to-run-11s/