AI-powered executive assistant skills that fully replace a human EA. Research meeting attendees, draft emails, create meeting briefs, and manage action items.

5 Skills

MIT License

Free Pricing

Installation

Open Claude Code and run this command:

/plugin install executive-assistant-skills@claude-code-plugins-plus

Use --global to install for all projects, or --project for current project only.

What It Does

AI-powered executive assistant skills for Claude Code that fully replace a human EA. These skills were born from a real experiment by Martin Gontovnikas (@mgonto), Auth0's former VP of Marketing/Growth, who discovered that a set of well-crafted Claude Code skills could handle everything his executive assistant used to do — meeting prep, email drafting, action item tracking, and daily digests.

The result: a suite of 5 skills that autonomously research attendees before meetings, draft personalized emails, extract and manage action items in Todoist, and deliver a morning executive briefing — all without human intervention.

Skills (5)

action-items-todoist SKILL.md View full skill →

Extract action items from today's Granola/Grain meetings, create Todoist.

ReadBash(gog:*)Bash(mcporter:*)Bash(todoist-cli:*)Bash(python3:*)

Action Items → Todoist + Email Drafts

Name: executive-assistant-skills
Author: Martin Gontovnikas

Config — read before starting

Read ../config/user.json (resolves to ~/executive-assistant-skills/config/user.json).

Extract and use throughout:

name, full_name — to identify your action items in meeting notes (e.g. "Gonto (Martin)")
whatsapp — for result delivery
workspace — absolute path to OpenClaw workspace

Do not proceed until you have these values.

Debug Logging (MANDATORY)

Read ../config/DEBUGLOGGING.md for the full convention. Use python3 {user.workspace}/scripts/skilllog.py action-items "" ['

'] at every key step. Log BEFORE and AFTER every external call (gog, mcporter, todoist-cli). On any error, log the full command and stderr before continuing.

Steps

0. Check today's calendar for meetings (BOTH accounts)

Before querying Granola, get today's actual meetings from BOTH calendars to know what to expect:


python3 {user.workspace}/scripts/skill_log.py action-items INFO "Starting action-items run"

# Get today's date in YYYY-MM-DD and tomorrow's
TODAY=$(date -u -d "$(TZ=America/Argentina/Buenos_Aires date +%Y-%m-%d)" +%Y-%m-%d)
TOMORROW=$(date -u -d "$(TZ=America/Argentina/Buenos_Aires date -d '+1 day' +%Y-%m-%d)" +%Y-%m-%d)

# Check BOTH calendars
gog --account {user.primary_email} --no-input calendar list primary --from "${TODAY}T00:00:00-03:00" --to "${TOMORROW}T00:00:00-03:00" --json 2>&1
gog --account {user.work_email} --no-input calendar list primary --from "${TODAY}T00:00:00-03:00" --to "${TOMORROW}T00:00:00-03:00" --json 2>&1

Log the results: python3 {user.workspace}/scripts/skill_log.py action-items DEBUG "Calendar events found" '{"primary": N, "work": M, "total": N+M}'

Merge events from both calendars. Filter for meetings with attendees (skip solo/personal events). This gives you the ground truth of what meetings happened today — use it to cross-check Granola results and catch any meetings Granola missed.

CRITICAL date syntax: Use explicit ISO8601 dates with -03:00 offset. Do NOT use relative expressions like +1 day, today, or tomorrow in gog flags — they may not be supported. Always compute the actual date strings.

1. Get today's meetings from Granola

Timezone note: Granola stores meeting times in UTC. For ART (UTC-3), querying "today" means using today's date AND tomorrow's date in UTC. E.g., for March 3 ART, query customstart: "2026-03-03" and customend: "2026-03-04" to capture all ART-day meetings.


mcport


                
                  
                  
                  email-drafting
                  SKILL.md
                  View full skill →
                
                
                  "Draft email replies for Gonto's Gmail accounts (m@gon.
                  
                      ReadBash(gog:*)Bash(mcporter:*)Bash(python3:*)GlobGrepWrite
                    
                
                
                  Email Drafting Skill
Config — read before starting
Read ../config/user.json (resolves to ~/executive-assistant-skills/config/user.json).
Extract and use throughout:

primaryemail, workemail — Gmail accounts
scheduling_cc — scheduling assistant email (CC on all scheduling emails, mention in body)
schedulingsilentcc — silent CC for scheduling visibility (do NOT mention in email body)
signature — sign-off for all drafts (e.g. "--gonto")
name — short name for context

Do not proceed until you have these values.
Debug Logging (MANDATORY)
Read ../config/DEBUGLOGGING.md for the full convention. Use python3 {user.workspace}/scripts/skilllog.py email-drafting  "" ['
'] at every key step. Log BEFORE and AFTER every external call (gog gmail, mcporter, todoist-cli). On any error, log the full command and stderr before continuing.
Overview
Auto-draft and manually-requested email drafts for {user.primaryemail} and {user.workemail}.
When to Use

Gmail hook detects a trigger (intro, scheduling, thanks/ack, positive reply)
User asks to draft/reply/send an email
Action items cron identifies email follow-ups needed

Architecture

Detects triggers and creates Gmail drafts
Handles all scheduling (slot-finding, conflict checking, calendar ops)
NEVER proposes specific dates/times or creates calendar events

Execution

"Reply" always means Reply All — include all original To + CC recipients. Only exclude if user explicitly says to reply to one person. Exception: intro handling moves the introducer to BCC per the intro sequence below.
After sending any email/draft, check if it fulfills an open Todoist task (send deck, intro, follow-up, etc.). If yes → complete the task immediately and confirm.
Hook/cron triggers: Always run via isolated sub-agent (prevents memory/context contamination)
Direct user requests ("draft a reply to X"): Main agent may execute directly, but must still follow all style rules and must NOT read MEMORY.md or daily memory files

Rules (non-negotiable)
Core

Draft-only mode — never send automatically
Mirror inbound language — match the language of the most recent non-automated message in the thread. If the thread has mixed languages, default to the language of the message you're replying to.
Always sign — end every draft with {user.signature}
Low confidence — don't draft; ask user for guidance


                
                  
                  
                  executive-digest
                  SKILL.md
                  View full skill →
                
                
                  "Generate the daily executive digest \u2014 a single WhatsApp summary\.
                  
                      ReadBash(gog:*)Bash(mcporter:*)Bash(todoist-cli:*)Bash(python3:*)
                    
                
                
                  Daily Executive Digest
Config — read before starting
Read ../config/user.json (resolves to ~/executive-assistant-skills/config/user.json).
Extract and use throughout:

primaryemail, workemail — both Gmail accounts to check
whatsapp — for delivery
workspace — absolute path to OpenClaw workspace

Do not proceed until you have these values.
Debug Logging (MANDATORY)
Read ../config/DEBUGLOGGING.md for the full convention. Use python3 {user.workspace}/scripts/skilllog.py exec-digest  "" ['
'] at every key step. Log BEFORE and AFTER every external call (gog, mcporter, todoist-cli). On any error, log the full command and stderr before continuing.
Steps
1. Read rules and state

Read {user.workspace}/style/DIGEST_RULES.md for format and rules
Read state files:
{user.workspace}/state/scheduling-threads.json — stalled threads (>3 days since proposed)
{user.workspace}/state/decisions-memory.json — context on people/companies
{user.workspace}/state/digest-state.json — avoid repeating items

Schema: {"lastRun": "ISO date", "surfacedItems": [{"id": "id|taskid>", "type": "intro|followup|draft|task|calendar", "surfacedAt": "ISO date"}]}. Use consistent IDs: Gmail thread IDs for email items, Todoist task IDs for tasks, calendar event IDs for calendar items. Do NOT use semantic strings like "calendar:golf-mar5" — always use the actual service ID. An item is "repeated" if its ID appeared in the last digest run. Re-surface only if its status changed since then.
Error handling
If any data source (Gmail, Calendar, Todoist, Granola) fails or times out:

Log the error, note it in the digest as "⚠️ [Source] unavailable — skipped"
Continue with remaining sources — never halt the entire digest for one failure
If BOTH Gmail accounts fail, abort and notify: "⚠️ Digest failed — Gmail unreachable"

2. Check Todoist

source {user.workspace}/.env
todoist-cli review

Include: overdue tasks, today's tasks, inbox (needs triage).
Format as "📋 Open Tasks" section: task name, due date, priority.
Highlight overdue first. Skip no-due-date tasks unless in inbox.
3. Check calendar (next 7 days) — BOTH accounts MANDATORY

gog --account {user.primary_email} --no-input calendar list primary --from "<today>T00:00:00-03:00" --to "<today+7>T00:00:00-03:00" --json
gog --account {user.work_email} --no-input calendar list primary --from "<today>T00:00:00-03:00" --to "<today+7>T00:00:00-03:00&q


                
                  
                  
                  meeting-prep
                  SKILL.md
                  View full skill →
                
                
                  "Prepare briefings for today's meetings \u2014 attendee research, email\.
                  
                      ReadBash(gog:*)Bash(mcporter:*)Bash(python3:*)Bash(openclaw:*)
                    
                
                
                  Daily Meeting Prep
Config — read before starting
Read ../config/user.json (resolves to ~/executive-assistant-skills/config/user.json).
Extract and use throughout:

name, full_name — user's name
primaryemail, workemail — Gmail accounts to check
whatsapp — WhatsApp number for delivery
timezone — IANA timezone (e.g. America/Argentina/Buenos_Aires)
slack_username — Slack DM target
workspace — absolute path to OpenClaw workspace (e.g. ~/.openclaw/workspace)

Do not proceed until you have these values.
Debug Logging (MANDATORY)
Read ../config/DEBUGLOGGING.md for the full convention. Use python3 {user.workspace}/scripts/skilllog.py meeting-prep  "" ['
'] at every key step. Log BEFORE and AFTER every external call (gog, mcporter, Granola, web search). On any error, log the full command and stderr before continuing.
Scope

Timezone: {user.timezone}
Calendars: primary of {user.primaryemail} AND {user.workemail}
Today's meetings only

Timezone note: Use explicit ART-bounded ISO8601 timestamps for calendar queries, NOT --date. Example: gog calendar list primary --account  --from "2026-03-03T00:00:00-03:00" --to "2026-03-04T00:00:00-03:00" --json. The --date flag uses UTC boundaries which misaligns with ART.

ALL meetings with attendees — both external and internal
Skip personal/solo events with no attendees (e.g. "Personal Trainer", "Golf", all-day reminders)

Meeting types
External meetings (attendees outside your email domains)
Full research brief (email context, Granola, LinkedIn, company research) — see below.
Internal meetings (all attendees from your email domains: hypergrowthpartners.com, growth.li)
Lighter brief — no LinkedIn/company research needed, but still include:

Attendee list
Granola context from previous instances of this recurring meeting
Recent email threads related to the meeting topic/agenda
Any open action items from last time
Format: same structure but skip LinkedIn/company sections

Recurring collaborative meetings (e.g. podcasts, content sessions with external co-hosts)
These are external meetings — give them full briefs. Don't skip recurring meetings just because they're familiar.
Error handling

If gog calendar fails for one account: continue with the other account, note "⚠️ [account] calendar unavailable" in output.
If Granola/Grain fails: continue without meeting history, note it per meeting.
If Wha


                
                  
                  
                  todoist-due-drafts
                  SKILL.md
                  View full skill →
                
                
                  Check Todoist for tasks due today (and overdue) that involve pinging,.
                  
                      ReadBash(gog:*)Bash(mcporter:*)Bash(todoist-cli:*)Bash(python3:*)
                    
                
                
                  Todoist Due-Today Email Drafts
Config — read before starting
Read ../config/user.json (resolves to ~/executive-assistant-skills/config/user.json).
Extract and use throughout:

primaryemail, workemail — Gmail accounts
whatsapp — for notification delivery
workspace — absolute path to OpenClaw workspace
signature — email signature

Debug Logging (MANDATORY)
Read ../config/DEBUGLOGGING.md for the full convention. Use python3 {user.workspace}/scripts/skilllog.py todoist-due-drafts  "" ['
'] at every key step. Log BEFORE and AFTER every external call (todoist-cli, gog, mcporter). On any error, log the full command and stderr before continuing.
Steps
1. Get today's due tasks from Todoist

source {user.workspace}/.env
todoist-cli today --json

Also check overdue tasks:

todoist-cli list --filter "overdue" --json

2. Identify email/ping tasks
Filter for tasks whose content matches outreach intent:

Keywords: ping, email, follow up, follow-up, send, reach out, text, message, intro, connect, check in, nudge, reply, respond, draft
Pattern: any task that implies sending a communication to a specific person

Skip tasks that are clearly NOT outreach (e.g. "review doc", "read report", "build proposal").
3. For each outreach task, draft the email
Read and follow ~/executive-assistant-skills/email-drafting/SKILL.md for all drafting rules.
For each task:

Identify the recipient: Extract person/company name from task content + description
Pull meeting context (if available): If the task description references a meeting name, date, or was created by the action-items cron (check for Granola citation links or meeting metadata in description), retrieve the transcript via Granola/Grain. If no meeting reference exists in the task, skip directly to step 3 — draft from email history and task content alone.


   # Find the meeting in Granola
   mcporter call granola list_meetings --args '{"time_range": "custom", "custom_start": "<meeting-date>", "custom_end": "<meeting-date+1>"}'
   # Query for what was discussed with this person
   mcporter call granola query_granola_meetings --args '{"query": "What did {user.name} discuss with <recipient> and wh



      
      

      
      

      
      

      
      
  Ready to use executive-assistant-skills?
  
    
    
        
        View on GitHub
      
  




      
      
          Related Plugins
          
            
                excel-analyst-pro
                Professional financial modeling toolkit for Claude Code with auto-invoked Skills and Excel MCP integration. Build DCF models, LBO analysis, variance reports, and pivot tables using natural language.
              
                brand-strategy-framework
                A 7-part brand strategy framework for building comprehensive brand foundations - the same methodology top agencies use with Fortune 500 clients.
              
                openbb-terminal
                Open-source investment research terminal integration - comprehensive equity analysis, crypto tracking, macro indicators, portfolio optimization, and AI-powered financial insights using OpenBB Platform
              
                promptbook
                Opt-in session analytics for Claude Code. After setup consent, tracks prompts, tokens, build time, and lines changed per session. Publishes shareable build cards on promptbook.gg. Strict consent gating — no data transmitted without explicit opt-in. Background submission after session end.
              
                wondelai-blue-ocean-strategy
                Blue Ocean Strategy framework for creating uncontested market space. Use the Strategy Canvas, Four Actions Framework (ERRC), and value innovation to break the value-cost trade-off.
              
                wondelai-contagious
                Word-of-mouth and virality framework using the STEPPS model (Social Currency, Triggers, Emotion, Public, Practical Value, Stories). Engineer products and content that people naturally share.
              
          
        

      
      
          Tags
          
            executive-assistantmeeting-prepemailproductivityautomationtodoist