extrabrainapp/extrabrain-skills

use-extrabrain

Manage ExtraBrain reference documents through the standalone extrabrain CLI.

소스 보기

사용 방법

  1. 설치 명령을 복사하고 자신의 에이전트 환경에 설치하기 전에 소스 패키지를 확인하세요.
  2. 필요한 도구, 인증 정보, 지원 환경은 원본 문서에서 확인하세요.
  3. 입력과 원하는 산출물을 설명하고 출처와 결과를 대조하세요.
npx skills add https://github.com/extrabrainapp/extrabrain-skills --skill use-extrabrain

시작 전 확인

  • 설치에는 모델 구독, API 한도, 파일 접근 권한이 포함되지 않습니다.
  • 별점 수와 설치 수는 활동 지표이며 독립적인 품질·보안 검증이 아닙니다.
원본 Skill 문서

원본 저장소의 제목, 예시, 코드, 표, 링크, 이미지를 유지해 표시합니다.

Use ExtraBrain

Use the first-party extrabrain CLI for reference documents, pairing, capability discovery, and updates. Session preparation, recording controls, profiles, history, settings, and desktop app updates require the ExtraBrain app UI.

Establish the CLI contract

  1. Check extrabrain --version and extrabrain --help for the installed command surface.

If the command is missing, use the installation instructions in the ExtraBrain CLI repository.

  1. Before document operations, run extrabrain --json capabilities. The JSON envelope is

{ "code": number, "data": object | null, "message": string }. On success, confirm data.apiVersion is v1, data.available is true, and data.capabilities advertises the required operation. Read references/documents.md for the mapping, commands, pairing scopes, and recovery steps.

  1. ExtraBrain must already be running with local document automation enabled. Discovery is

public and does not prove the CLI is paired. If a document command reports an authentication or scope error, pair with the scopes needed for the user's task and have the user approve the connection in the app. Credentials are stored by the CLI in macOS Keychain.

Use --json for command results. The exit code and per-file statuses determine success; a partial import is not a completed import.

Work with documents

  • List existing documents before creating another copy. Use returned document IDs, revisions,

and index generations for later operations.

  • Import local files directly through the CLI. It must run in the same filesystem namespace

as the selected files; a remote shell or container cannot use a desktop-only path.

  • Search for relevant snippets before requesting extracted text. Read bounded text pages only

when needed. Use original export when the user needs the exact source bytes.

  • After an import, check its per-file results and batch/item status, then verify document

metadata and indexing status. After deletion, verify the document is absent from the list.

  • Delete only with explicit user intent, one known document ID, and its current revision.

On a revision conflict, reread metadata and confirm the changed target before retrying.

  • Let the CLI manage import and deletion retry identifiers. Resume only with the returned

resume ID for the same unchanged source files and import intent.

Keep external research and source gathering in your own authorized tools. Do not access ExtraBrain's database or managed source directory, extract credentials, or build a separate document extractor to bypass the CLI.

If the CLI or app API is unavailable, explain the reported limitation. For file imports, the user can use Settings > Personalization > Reference Documents and the native file picker.

Update the CLI

When the user asks to update the CLI, run extrabrain update, then check extrabrain --version. This updates the standalone command. Use the app UI for desktop application updates.

관련 작업