Todoist Due Drafts logo

Todoist Due Drafts

Community
mgonto
todoist-due-drafts

Check Todoist for tasks due today (and overdue) that involve pinging, emailing, or following up with someone. Auto-draft the emails using meeting context and notify via WhatsApp. Use when running the daily due-drafts cron, or when user asks to process email tasks from Todoist.

Overview

Publishermgonto
Repositoryexecutive-assistant-skills
Skill nametodoist-due-drafts
Stars
121
Forks
14
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by mgonto on GitHub. Read the source before you install it.

Installation

Install the Todoist Due Drafts AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/mgonto/executive-assistant-skills/tree/main/todoist-due-drafts
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/mgonto/executive-assistant-skills.git /tmp/executive-assistant-skills
mkdir -p .claude/skills
cp -r /tmp/executive-assistant-skills/todoist-due-drafts .claude/skills/todoist-due-drafts
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Todoist Due Drafts in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Todoist Due Drafts on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Todoist Due Drafts is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

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:

  • primary_email, work_email — Gmail accounts
  • whatsapp — for notification delivery
  • workspace — absolute path to OpenClaw workspace
  • signature — email signature

Debug Logging (MANDATORY)

Read ../config/DEBUG_LOGGING.md for the full convention. Use python3 {user.workspace}/scripts/skill_log.py todoist-due-drafts <level> "<message>" ['<details>'] 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

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

Also check overdue tasks:

bash
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:

  1. Identify the recipient: Extract person/company name from task content + description
  2. 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.
    bash
    # 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 what did he promise or commit to do?", "document_ids": ["<meeting_id>"]}'
    Then cross-check with Grain for the full transcript:
    bash
    mcporter call grain.list_attended_meetings --args '{}'
    # Note: Grain's schema does not support `start_date`/`end_date` filters. Call with empty args and filter the results manually by `start_datetime` to match the meeting date range.
    mcporter call grain.fetch_meeting_transcript --args '{"meeting_id": "<grain_meeting_id>"}'
    When meeting context is available, the email draft must reflect what was actually said — not just the task title. Look for: specific commitments, timelines discussed, names/projects mentioned, tone of the conversation, and any docs/links promised.
  3. Search email history: Find the latest thread with this person across both Gmail accounts to get context, their email address, and the right account to reply from. If no prior thread exists with this recipient, default to {user.work_email} for professional/business contacts or {user.primary_email} for personal contacts. When ambiguous, use {user.work_email}.
  4. Determine email type: follow-up, ping/check-in, intro, send-doc, etc.
  5. Draft the email: Create a Gmail draft on the correct account (the one with the existing thread). If it's a reply, use --thread-id to keep it in the same thread.
  6. Use all context together: Combine the meeting transcript (what was actually discussed), task description, and email history to write a specific, contextual draft. Never write generic follow-ups — reference concrete topics from the conversation.
Draft rules
  • Mirror the language of the existing thread (English or Spanish)
  • Keep it short — these are follow-ups and pings, not essays
  • Sign with {user.signature}
  • If the task says "ping" or "check in" — write a brief, friendly nudge
  • If the task says "send [thing]" — draft the email with the attachment reference (or attach if path is known)
  • If the task says "intro" — follow intro format from email-drafting skill
  • If recipient email can't be found — report it, don't skip silently

4. Notify via WhatsApp

Send a single WhatsApp message to {user.whatsapp} with:

📬 *Due-today drafts*

<N> email drafts created from today's Todoist tasks:

1. **<recipient>** — <one-line intent> (draft in <account>)
2. ...

Review and send when ready.

If tasks exist but none are outreach-type, report:

📋 *Due-today tasks (no drafts needed)*

<list of today's tasks — quick reference>

If no tasks due today → NO_REPLY (skip notification entirely).

5. Infrastructure

bash
python3 {user.workspace}/scripts/cron_canary.py ping todoist-due-drafts

Error handling

  • If todoist-cli fails: notify via WhatsApp "⚠️ Todoist due-drafts failed: ", ping canary, exit.
  • If Granola/Grain lookup fails for a task: continue without meeting context — draft from email history + task content only.
  • If Gmail draft creation fails for one task: continue with remaining tasks, report failures in the notification.
  • If no .env file or token expired: report "⚠️ Todoist token unavailable — check agent-secrets lease" and exit.

Rules

  • Don't complete the tasks — just draft and notify. User decides when to send.
  • Before drafting, check if user already replied: gog --account <account> --no-input gmail search "to:<recipient> in:sent newer_than:14d" --json. If recent sent mail exists in the same thread, skip drafting and note "already replied" in the notification.
  • Check for existing drafts before creating: gog --account <account> --no-input gmail drafts --json. If a draft already exists for the same thread (matching thread ID or recipient + subject), skip and note "draft already exists."
  • Overdue outreach tasks get a ⚠️ prefix in the notification.

Frequently asked questions

What does the Todoist Due Drafts AI skill do?

Check Todoist for tasks due today (and overdue) that involve pinging, emailing, or following up with someone. Auto-draft the emails using meeting context and notify via WhatsApp. Use when running the daily due-drafts cron, or when user asks to process email tasks from Todoist.

Why use Todoist Due Drafts on TypingMind?

Because you install it once and use it with any model. Todoist Due Drafts is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Todoist Due Drafts in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/mgonto/executive-assistant-skills/tree/main/todoist-due-drafts. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Todoist Due Drafts?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Todoist Due Drafts?

As many as you like. As long as a model supports skills, you can use Todoist Due Drafts with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Todoist Due Drafts AI skill free?

It is published on GitHub by mgonto. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇