Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.asteragents.com/llms.txt

Use this file to discover all available pages before exploring further.

​
What it does

The Manage Apps tool lets agents create and manage persistent React applications that render inside Aster at /apps/{id}. Agents write TSX source code, provide JSON snapshot data, and the platform compiles and hosts the result. Users can open app URLs anytime to see interactive dashboards, reports, and data views. See Apps for the full feature overview.

​
Parameters

ParameterTypeRequiredDescription
actionstringYes"create", "list", "get", "update", "delete", "load_to_sandbox", or "publish_from_sandbox"
app_idintegerVariesRequired for get, update, delete, load_to_sandbox. Optional for publish_from_sandbox (omit to create new, provide to update existing)
app_dirstringVariesRequired for publish_from_sandbox. Optional for load_to_sandbox (defaults to /home/user/aster-apps/app-{app_id}). Must be under /home/user/aster-apps/
namestringFor createDisplay name of the app
descriptionstringNoHuman-readable description
source_tsxstringFor create/updateFull React TSX source code. Must export a default component
snapshot_dataobjectNoJSON data the app renders, accessed via useAppData() hook

​
Actions

​
create

Creates a new app from TSX source and snapshot data in a single call. Required: name, source_tsx
Optional: description, snapshot_data
Returns: { id, name, url }

​
list

Returns all apps in the organization. Returns: { apps: [{ id, name, description, agentId, createdAt, updatedAt }] }

​
get

Returns full details for a single app, including the current version’s source code. Required: app_id
Returns: { id, name, description, url, currentVersion: { id, sourceTsx, bundleHash, createdAt } }

​
update

Updates an existing app’s source, data, or metadata. Creates a new internal version. Required: app_id
Optional: name, description, source_tsx, snapshot_data

​
delete

Soft-deletes an app. Required: app_id

​
load_to_sandbox

Hydrates an existing app into the sandbox for editing. The agent can then use execute_python to inspect and modify files before re-publishing. Required: app_id
Optional: app_dir (defaults to /home/user/aster-apps/app-{app_id})
Returns: { app_id, app_dir, version_id, files } Creates this directory structure in the sandbox: 
/home/user/aster-apps/app-{id}/
  aster-app.json
  src/App.tsx
  data/snapshot.json
  assets/           (if the app has asset files)

​
publish_from_sandbox

Reads files from the sandbox, validates the manifest, compiles the TSX, and publishes to Aster. Required: app_dir
Optional: app_id (omit to create a new app, provide to update an existing one)
Returns: { id, name, url, version_id, source_files, asset_files } The directory must contain a valid aster-app.json manifest. See Apps > The aster-app.json Manifest.

​
Available imports in app source

ImportProvides
reactReact core
@aster/sdkuseAppData() hook
@aster/uiShadcn components (Card, Badge, Button, Table, Tabs, Progress, Separator, ScrollArea, Tooltip) + cn()
rechartsCharts (LineChart, BarChart, PieChart, AreaChart, ComposedChart, etc.)
lucide-reactIcons
date-fnsDate utilities

​
Common use cases

​
Analytics dashboard from live data

“Pull our GA4 traffic data for the last 90 days and build me a dashboard showing trends by channel, top pages, and device breakdown” The agent queries GA4, shapes the data, and publishes an interactive dashboard with charts and tables.

​
CRM pipeline report

“Build a pipeline dashboard from Salesforce showing open deals by stage, weekly trends, and top opportunities” The agent queries Salesforce, computes KPIs, and creates a persistent view the sales team can bookmark.

​
Updating an existing app

“Update app 5 with the latest data” The agent re-runs its data-gathering tools, then calls update with the new snapshot data (and optionally updated TSX).

​
Complex multi-file app via sandbox

“Build a detailed analytics app with separate components for each section — overview, queries, pages, and opportunities” The agent uses execute_python to write multiple TSX files and a data-processing script, then publishes from the sandbox.