TreeMap
무료 오픈소스 디스크 사용량 시각화 도구로, 사각형 트리맵·중복 파일 탐색·디스크 가득 예측·정리 시 파일을 휴지통으로만 이동한다
README 전체를 불러오는 중입니다. 아래는 저장된 발췌입니다.
Helio turns raw data into dashboards you can build by asking.
Point it at a CSV, a SQL database, or a REST API. Shape the data with a visual
pipeline. Place any node's tail as an Output — a table, a metric, a chart — and
arrange them on a responsive grid. Then keep going in plain language: describe a
change, review the exact diff, and apply it.
Describe a change, review it, apply it
Helio's assistant edits real dashboards. It never applies anything silently: every
request becomes a patch set you see as a before/after diff, accept or reject,
and undo afterwards.README 전문 보기 ↗Helio turns raw data into dashboards you can build by asking.
Point it at a CSV, a SQL database, or a REST API. Shape the data with a visual pipeline. Place any node's tail as an Output — a table, a metric, a chart — and arrange them on a responsive grid. Then keep going in plain language: describe a change, review the exact diff, and apply it.
Helio's assistant edits real dashboards. It never applies anything silently: every request becomes a patch set you see as a before/after diff, accept or reject, and undo afterwards.
Everything in Helio follows one path, and every panel is a placement of an Output:
Data source → Pipeline → Outputs → Dashboard
CSV filter table panel
SQL aggregate metric panel
REST date bucket chart panel
files join, pivot ... panel
A pipeline is an ordered list of transformation steps over one source. Twenty-three operations ship today — filter, aggregate, join, pivot, window, unpivot, date bucket, lookup, dedupe, and more — each previewable step by step, with the output schema computed before you run it.
An Output places a live tail of the pipeline's node tree — the whole source, or any intermediate step — as a table, metric, chart, or other kind. A dashboard panel is a placement of one Output; because binding always goes through an Output, a panel can never silently drift from the shape of its data.
The dashboard below is built from Helio's own delivery record — 459 merged pull requests and 805 Linear tickets — ingested as two CSVs and shaped by five pipelines.
Helio is designed to be driven by agents as well as people. The same operations the
UI performs are available over an MCP server (helio-mcp/), authenticated with
a personal access token — list sources, author pipelines, create and bind panels,
propose a whole dashboard, apply a patch set, undo it.
The delivery-analytics dashboard above was built end to end through that server.
| Local setup, database, environment | docs/cloud-dev-setup.md |
| Architecture, commands, conventions | CLAUDE.md |
| Deployment and infrastructure | docs/deployment.md |
| Design language and UI standards | DESIGN.md |
| Contributing and code quality | CONTRIBUTING.md |
| API contracts | schemas/ and openspec/ |
React 19 · TypeScript · Redux Toolkit · Vite — frontend Scala 2.13 · Apache Pekko HTTP · Slick · PostgreSQL — backend Flyway migrations · JSON Schema 2020-12 contracts · Cloud Run + Firebase Hosting
Releases are cut with /release <major.minor>, which tags the release branch and
publishes a changelog. Pushing a v* tag is what deploys.
See LICENSE.