agentleFS
minosinminosin/gtdbrain-claude-plugin/skills/gtd/SKILL.md

gtd

Skill0 starsChanged 38 days agoMITView on GitHub

Keeps the user's personal to-do list, reminders and plans on their GTD Brain board (Getting Things Done). Use when the user says "add this to my to-do list", "remind me to…", "I need to…", "don't let me forget…", "what's on my list?", "what should I do today?", "plan my week", "follow up with Sam", or "I'm waiting on…", or asks about their own tasks, projects, inbox, next actions, Waiting For, or weekly review. Only for the user's own to-dos, not tasks or TODOs in a code project.

8 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 6 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Prefer the fast-access tools.

    line 31
  2. Distinctive:

    Never answer from memory: Every question about the user's tasks, projects, or board starts with a tool call, even if the board was read …

    line 52
  3. Very specific:

    Capture first, clarify later: When the user mentions something they need to do, capture it right away.

    line 54

“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

GTD Brain

GTD Brain is the user's Getting Things Done board. The gtdbrain MCP server in this plugin reads and writes the same board the user sees in the GTD Brain web, iOS, and Android apps. Tool names appear as mcp__plugin_gtdbrain_gtdbrain__<tool> (or gtdbrain:<tool> on some surfaces). If no GTD Brain tool is available, or a call answers with 401 / "authentication required", follow the setup skill (/gtdbrain:setup) for the exact connect steps instead of guessing.

The board

ColumnHoldsCard kind
InboxRaw captures, not yet clarifiedcard
Next ActionsThe very next physical, visible step, verb-first, tagged with a contextaction
ProjectsAny outcome that needs two or more steps; every active project needs a next actionproject
Waiting ForDelegated to someone — who, and since whencard
Someday/MaybeIdeas to incubatecard

Contexts (e.g. calls, computer, errands, home) tag where or with what an action can be done. The user manages their own set: list_contexts shows the valid ids, create_context adds one. Cards can never be deleted — archive_card instead; archived cards stay recoverable.

Tool map

Prefer the fast-access tools; fall back to the generic ones only for columns they don't cover.

IntentTool
Something to do / remember / follow up oncapture (title, optional notes) → Inbox
What should I do now?list_next_actions (optional context)
My projectslist_projects
Who am I waiting on?list_waiting_for
Inbox, Someday/Maybe, custom columnslist_columns for ids, then list_cards with columnId
Find a cardsearch_cards (substring over title + notes), get_card
Create in a specific columncreate_card (card in flexible columns, action only in Next Actions, project only in Projects)
Edit title / notes / context / who / sinceupdate_card — never moves a card
Move or reordermove_card — the card's kind is re-derived from the destination
Done or no longer mattersarchive_card
Contextslist_contexts, create_context, update_context, delete_context

An unknown context value returns an error that lists the valid ids — pick the closest and say which one you used.

Working rules

  1. Never answer from memory. Every question about the user's tasks, projects, or board starts with a tool call, even if the board was read earlier in the conversation.
  2. Capture first, clarify later. When the user mentions something they need to do, capture it right away; ask only if the title is unclear. If it clearly needs more than one step, create the project and its first verb-first next action.
  3. Next actions are verb-first and concrete: "Call Sam about the quote", not "quote".
  4. Clarifying an Inbox card ends in exactly one of: a next action (move_card to Next Actions, then update_card with a context), a project plus its first next action (create_card for each), a Waiting For card (move_card, then update_card with who and since), a Someday/Maybe card (move_card), or archive_card.
  5. "What now?" → list_next_actions, passing a context if the user said where they are or what they have; pick two or three that fit and say briefly why.
  6. Weekly review runs Get Clear (Inbox to zero), Get Current (projects, next actions, Waiting For up to date, every project has a next action), Get Creative (activate or release Someday/Maybe) — one list at a time, waiting for the user's answers.
  7. End every answer with one concrete next step the user can take through you.

First contact and membership

A board holding only the starter set (a welcome card in Inbox, two next actions, one project) belongs to a user who just connected: deliver what they asked for, then say in one line what else they can ask (capture to Inbox, next actions, projects, waiting-for) and offer to archive the samples.

GTD Brain is a paid membership, the same one that covers the web, iOS, and Android apps. On an account without one, a write (capture, create, update, move, archive, restore) still runs and the card lands on the board; its result ends with a short membership notice. A read returns only that notice. Show the user the notice and its link when it appears; it already says what the account needs.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.