Brief
Capture problems and ideas as structured documents with intent, facts, and questions.
Installation
- Make sure Claude is on your device and in your terminal.
Skills load from
~/.claude/skills/when Claude Code starts up — so you need it on your machine first. If you don't have it yet, install it once with the command below, then runclaudein any terminal to verify.One-time setupnpm i -g @anthropic-ai/claude-codeAlready have it? Skip ahead.
- Paste into Claude Code or into your terminal.
This copies the whole skill folder into
~/.claude/skills/brief-greenfield-inc/— the SKILL.md plus any scripts, reference docs, or templates the skill ships with. Works for every skill.What this command does
- Makes a new, empty temporary folder.
- Downloads the skill's folder,
main/src/services/paneChatBundle/skills/brief, fromgithub.com/greenfield-inc/Pane(branchmain), the GitHub repository it's listed from. You get the version that is there now, not a copy we keep. - Copies that folder to
~/.claude/skills/brief-greenfield-inc/. - Deletes the temporary folder.
Installing runs none of the skill's files; it only downloads and copies them. When you later use the skill, Claude follows its SKILL.md and may run scripts that come with it. Read them first.
Faster alternative (instruction-only skills)
Skips the clone and grabs only the SKILL.md file. Don't use this if the skill ships Python scripts, reference markdowns, or asset templates — they won't be downloaded and the skill will fail when it tries to load them.
Quick install (SKILL.md only)Sign up to copy - Restart Claude Code.
Quit and reopen Claude Code (or any other agent that loads from
~/.claude/skills/). New skills are picked up on startup. - Just ask Claude.
Skills auto-activate when your request matches the skill's description — no slash command needed. Trigger phrases live in the skill's own frontmatter; you can read them in the “What this skill does” section above.
Prefer to read the source first? Open on GitHub.
When Claude uses it
Use when a problem, idea, or change of intent is worth capturing as a long-form page, or the person asks to write it up. Never chooses a solution; the tracker issue belongs to create-ticket.
What this skill does
Brief
A brief says what is wrong or wanted, for whom, and how a user would know it is solved. It survives a change of approach, because it contains no approach. Alternatives belong in the options document. Steps belong in the plan.
Capture the intent
- Preserve the person's language and constraints. Ground every statement in a source: the conversation, an issue, a bug report, the code.
- Keep four things visibly apart: facts, decisions already made, proposals, and open questions. Label suspected causes and unmeasured impact as such.
- Capture the trigger: what happened, what was expected, and what it costs. For a new idea, capture the unmet need.
- When the request arrives with a solution attached ("add patches to cut editing costs"), look at the actual behaviour and cost first, and write the problem underneath it.
- Ask what is missing instead of inventing it: what prompted this, who is affected, what a good result would enable, what would make the current approach good enough.
- Read the code, or give a fresh-context subagent one question to answer with the
gather-evidenceskill, when a fact about today's behaviour is needed. Give the subagent just that question. - Research settles behaviour, never approach. A researched answer about what a user observes goes in as a labelled proposal with its source, pending the person's confirmation; anything about how it is built is only named here and argued in the options document.
When intent changes
Revise the same brief. Mark the earlier goal as superseded, cite the decision that replaced it, refresh success and constraints, and add a change-log line. Keep constraints that still hold. If what changed is only the approach, the brief does not change: that belongs in the options document.
One brief or several
Keep one coherent outcome together. Split work that has independent outcomes, owners, or release timing, and link the shared context.
Layout
An HTML page for a person, rendered with the page house standard. Layout: references/brief-layout.md. Someone who was not in the conversation should understand the problem in a minute.
Save and link
Draft freely during discussion. Save or publish only when the person asks or confirms it is ready. Filing an issue does not mean work starts. When you save, follow references/saving.md: the bundle first, the tracker issue when they use one, then read back what you saved.
Ready check
Before calling it ready, read it as the person who will pick it up: can they say what is wrong, why it matters, how success is observed, and which decisions are still open? Set the status chip accordingly. When the person is ready to choose a direction, the brief goes to Socrates review with the options document, as the options skill describes.
Banned
- An architecture or implementation presented as decided
- Acceptance criteria written as implementation steps
- Implementation steps, file lists, or a plan
Related skills
Documentation Co-Authoring
anthropics
Guide structured workflows for writing docs, proposals, and technical specs collaboratively.
MCP Server Builder
anthropics
Build protocol servers that connect language models to external APIs and services.
Ask Questions If Underspecified
trailofbits
Ask clarifying questions before starting work on ambiguous requests.
Claude Plugin Marketplace Builder
daymade
Convert a Claude skills repository into an installable marketplace with auto-updates.