agent-browser
Browser automation using Vercel's agent-browser CLI. Use when you need to interact with web pages, fill forms, take screenshots,…
Connect Todoist so Dex reads and updates your Todoist tasks two ways. Use when the user says 'I use Todoist', 'sync Todoist', or pastes a todoist.com link. Not for Things 3 (`things-setup`) or Trello (`trello-setup`); not for Jira tickets (`atlassian-setup`).
Claude Code
$ npx -y skills add davekilleen/Dex --skill todoist-setup --agent claude-code
Codex
$ npx skills add davekilleen/Dex --skill todoist-setup -a codex -y
Cursor
$ npx skills add davekilleen/Dex --skill todoist-setup -a cursor -y
Gemini CLI
$ npx skills add davekilleen/Dex --skill todoist-setup -a gemini-cli -y
GitHub Copilot
$ npx skills add davekilleen/Dex --skill todoist-setup -a github-copilot -y
OpenCode
$ npx skills add davekilleen/Dex --skill todoist-setup -a opencode -y
Windsurf
$ npx skills add davekilleen/Dex --skill todoist-setup -a windsurf -y
name: todoist-setup
description: "Connect Todoist so Dex reads and updates your Todoist tasks two ways. Use when the user says 'I use Todoist', 'sync Todoist', or pastes a todoist.com link. Not for Things 3 (`things-setup`) or Trello (`trello-setup`); not for Jira tickets (`atlassian-setup`)."
integration:
id: todoist
name: Todoist
mcp_server: null
auth: api_key
enhances:
- skill: daily-plan
capability: "Can pull in your Todoist tasks due today when you ask"
- skill: triage
capability: "Can route an item to Todoist instead of Dex when you ask"
new_capabilities:
- name: On-request Todoist access
trigger: "Ask Dex to check, create, or complete Todoist tasks in any conversation"Connect your Todoist account so Dex can work with your Todoist tasks whenever you ask — check what's due, create a task there instead of in Dex, or mark something done.
**What this enables (on request):** Once connected, say "sync with Todoist" — Dex pushes new tasks to the right Todoist project, marks completions, and pulls back tasks you completed or created in Todoist directly. Sync runs on demand; Dex does not poll Todoist in the background without you asking.
Once connected, you can ask Dex to:
---
1. Check `System/integrations/config.yaml` for a `todoist:` section with `enabled: true` through `core.utils.strict_yaml.load_yaml_path`; refuse duplicate keys, aliases, anchors, or merge mappings before reading or writing any setup field. 2. If enabled, test authentication with the read-only health check 3. If healthy, skip to **Reconfiguration** section below 4. If not configured or unhealthy, continue to Step 2
Say:
**Let's connect Todoist to Dex.** Once connected, you can ask me to check, create, or complete Todoist tasks right from our conversations. **What you'll need:** - Your Todoist API token (I'll show you where to find it) - About 2 minutes **Ready to go?**
Wait for confirmation.
Guide the user:
To get your Todoist API token: 1. Open Todoist (web or app) 2. Go to **Settings** → **Integrations** → **Developer** 3. Copy the **API token** shown there Do not paste the token into this conversation. Open the ignored vault-root `.env` in your local editor and replace the placeholder in this line directly: `TODOIST_API_KEY=<paste token locally here>` Save the file with mode `0600`, then reply only `saved`. Dex must never echo, read aloud, log, or include the value in a command, argv, or process environment.
Wait only for the non-secret `saved` confirmation. Never ask the user to provide or validate the token in chat.
Confirm locally that `.env` defines `TODOIST_API_KEY` without printing or returning its value. Use `python3 -m core.utils.credential_workflow scan` for the redacted authority finding; repair permissions/ownership before continuing if it reports an invalid `.env` authority. Preserve unrelated lines and never place the value in `.mcp.json`, tracked YAML, a command, argv, logs, transcript, or process environment. Existing `.mcp.json` is scan/report-only and must remain byte-identical.
Before health, update only the non-secret tracked Todoist fields to `enabled: true` and `api_key_env_var: TODOIST_API_KEY`; do not add an `api_key` field.
Use only Dex's sanitized Python-to-adapter-stdin read-only health path. It resolves `.env` internally and performs Todoist `GET /projects`; it does not put the token in the command or environment:
python3 -c 'from core.integrations.task_sync import check_service_health; print(check_service_health("todoist"))'**If the health result is `{"healthy": True}`:**
Connected! Todoist authentication succeeded through the read-only health check.
**If it fails:**
That API key didn't work. A few things to check: 1. **Copy the full key** — it should be about 40 characters 2. **Check for extra spaces** before or after the key 3. **Regenerate the key** in Todoist Settings → Integrations → Developer Want to try again?
Retry up to 2 times, then offer to skip and come back later.
Ask the user which Todoist project should receive Dex tasks:
**Which Todoist project should Dex tasks go into?** The health check does not enumerate projects. Open Todoist locally and enter the exact project name you want Dex to use without pasting any credential. You can pick one default project, or map each Dex pillar to a different project. **Option A:** All Dex tasks go to one project (simplest) **Option B:** Map each pillar to a project: (Read pillar names from `System/pillars.yaml` and list them here) - [pillar 1 name] → [project] - [pillar 2 name] → [project] - [pillar 3 name] → [project] Which works for you?
Save their choices for the config file.
Write to `System/integrations/config.yaml` — update the todoist section. Build `pillar_map` dynamically from the user's actual pillars in `System/pillars.yaml` (on
One personal Dex, built around your work. Keep your priorities, people, meeting notes and tasks in a folder you own. Work with Dex through your AI app to plan the day, prepare for a meeting or follow through on a promise.
Browser automation using Vercel's agent-browser CLI. Use when you need to interact with web pages, fill forms, take screenshots,…
Build applications where agents are first-class citizens. Use this skill when designing autonomous agents, creating MCP tools,…
This skill should be used when writing Ruby gems following Andrew Kane's proven patterns and philosophy. It applies when creating…
This skill should be used before implementing features, building components, or making changes. It guides exploring user intent,…
Capture solved problems as categorized documentation with YAML frontmatter for fast lookup
Expert guidance for creating, writing, and refining Claude Code Skills. Use when working with SKILL.md files, authoring new…
Shape output for a reader with ADHD: lead with the next action, number multi-step work, restate state across turns, suppress…
Set up the owner profile so a personal agent knows who it works for (first-run setup, \"get to know me\"), or distill colleague,…