plannotator-setup-goal Skill for Claude Code: Turn an Idea Into a Goal

Turns a rough idea into a goal package for /goal: an interview, a reviewed fact sheet, a gated plan and a goal file.

Installs: 12K; on skills.sh
Stars: 9.8K; on GitHub
Files: 2; besides SKILL.md
Updated: Nov 2026; last repo commit

plannotator-setup-goal makes Claude turn an idea into a folder at goals/<slug>/. It restates your goal, asks questions in a Plannotator browser form, turns your answers into a fact sheet you accept or edit, explores the codebase to write a plan, and waits for you to approve that plan. It ends with goals/<slug>/goal.md, which you launch with /goal goals/<slug>/goal.md. It is set to not start on its own, so you call it by name.

It needs the plannotator command on your machine, because the interview, fact review and plan gate are Plannotator commands that open browser sessions. It writes JSON and markdown files into your project's goals/ folder. For a vague goal it can first run an optional one-question-at-a-time grilling pass, adapted from Matt Pocock's grill-me skill.

Good for

Developers who want a vague objective pinned down before an agent runs long on it
Plannotator users who already review plans in the browser
Teams that want written facts and a plan to check a goal against

Install plannotator-setup-goal

Claude Code

Copy this prompt.

Install the plannotator-setup-goal skill in Claude Code. Run this command:

curl -fsSL https://plannotator.ai/install.sh | bash

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Requires bash.

The skill runs the plannotator command, so install the binary first. The installer adds hooks and slash commands for the agents it detects; add --minimal for the binary only. The skills CLI command installs the three extra skills in that folder (compound, setup-goal and visual-explainer). Call the skill by name; it never starts on its own.

From the project README. Source.

What's in the skill

When it loads

Turn an idea or objective into a goal package for /goal. Interviews the user, builds a reviewed fact sheet via Plannotator, then explores the codebase to produce an execution plan.

Sections in SKILL.md

Phases
1. Rearticulate
2. Interview Bundle
3. Fact Sheet
4. Plan
5. Goal Output

Key facts

Author: backnotprop (Plannotator repo owner)
License: Apache 2.0 or MIT, at your option (repo README)
Installs on skills.sh: 11,967 (Nov 2, 2026)
GitHub stars: 9,810 (repo, Nov 2, 2026)
Files: 3 including SKILL.md: a Codex metadata file and a test file, no scripts
Needs: The plannotator command, run in a project folder
Starts: Only when you call it (disable-model-invocation)
Last commit: Nov 1, 2026 (repo)

What it includes

States your goal back in 2-3 sentences and creates goals/<slug>/.
If the goal is vague it does shallow codebase exploration first, then waits for you to confirm or correct before continuing.
Runs code · 1. Rearticulate
A one-question-at-a-time interview for vague or interdependent goals.
Suggested for vague goals and run when you ask. The protocol is taken verbatim from Matt Pocock's grill-me skill, as the SKILL.md comment says.
Optional grill-first pass
Writes interview.json and runs plannotator setup-goal interview.
A browser form of questions with recommended answers. The agent waits for you to submit and saves the result to interview-result.json. It does not ask obvious questions.
Runs code · 2. Interview bundle
Builds facts you accept, edit or remove, with automated checks preselected.
Runs plannotator setup-goal facts. The result is saved as facts-result.json, facts.md and facts.meta.json.
Runs code · 3. Fact sheet
Explores the codebase and writes plan.md, gated by plannotator annotate --gate.
Ordered steps, verification for each, and risks. If you deny it, the agent revises and gates again until approved.
Runs code · 4. Plan
Writes goal.md and tells you how to launch it with /goal.
The goal in 1-3 sentences, links to facts.md and plan.md, and a done condition.
Write · 5. Goal output
Codex display name and a default prompt using $plannotator-setup-goal.
Metadata for Codex, shown as Setup Goal.
agents/openai.yaml

From the official docs

Permissions

Medium risk

It runs the plannotator command, writes files into a goals folder in your project, and reads your codebase. Run it in the project you mean to plan and review the plan before approving.

What it can do

Read your data
Create and change things
Delete things
Read-only mode

Watch out for

The first install route pipes a script into bash
The README installer is curl -fsSL https://plannotator.ai/install.sh | bash. The README says a full install downloads the binary from GitHub Releases and can also contact GitHub, Ataraxy-Labs/sem and npm. Read the script first, or pass --minimal for the binary only. github.com
It writes files into your project
It creates goals/<slug>/ with interview, facts, plan and goal files. Add the folder to version control or delete it as you prefer. github.com
It needs the plannotator command
The skill runs plannotator setup-goal and plannotator annotate. The Plannotator installer also sets up hooks and slash commands for detected agents, so pass --minimal if you want the binary only. github.com

Technical details

Scopes	plannotator project goals/ folder

Questions

What does plannotator-setup-goal do?

It turns a rough idea into a goal package in goals/<slug>/. You answer a browser interview, accept a fact sheet, approve a plan, and get a goal.md to launch with /goal.

How do I install plannotator-setup-goal?

Use the install box above. You need the skill and the plannotator command. The skills CLI command installs the skill, and the Plannotator installer adds the binary.

Does it need Plannotator?

Yes. The interview, fact review and plan gate are plannotator setup-goal and plannotator annotate --gate commands that open browser sessions.

How do I start it?

Call it by name. Its frontmatter sets disable-model-invocation, so Claude does not start it on its own. The Codex metadata uses $plannotator-setup-goal.

Is it safe to run?

It writes files into a goals folder in your project and runs the plannotator command. The Plannotator installer also configures hooks for agents it detects, and --minimal skips that. Review the plan before you approve it.

How is it different from grill-me?

This skill can run the grill-me interview protocol as an optional first step, then continues into a fact sheet, a gated plan and a goal file. The SKILL.md copies that protocol and credits Matt Pocock.

Information

updated Nov 2, 2026

Repository: backnotprop/plannotator
Folder: apps/skills/extra/plannotator-setup-goal

backnotprop/plannotator skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 6

Plannotator README · github.com
Plannotator install.sh · github.com
plannotator-setup-goal SKILL.md · github.com
plannotator-setup-goal on skills.sh · skills.sh
backnotprop/plannotator · github.com
openai.yaml · github.com