SDK Platform Coverage
CredVault is a data platform with a dashboard, REST APIs, developer SDKs, CLI tooling, and connected services. The SDKs are for developers who want to call CredVault from their own code. They are not a replacement for every screen in the dashboard.
Use this page to decide whether a feature should be used through the dashboard, REST API, SDK, or CLI.
How The SDK Fits
The SDK wraps CredVault API calls into language-native methods. Instead of manually writing every HTTP request, a developer can create a client and call resources such as data, cie, functions, webhooks, or logs.
The SDK is most useful when you are building:
- backend services that read and write CredVault data
- data pipelines that push records into CredVault
- automation jobs that create backups or inspect logs
- AI and analytics workflows that call CIE models
- internal tools that manage functions, triggers, and webhooks
- scripts that run from CI/CD, notebooks, or worker machines
The SDK is not meant to reproduce the full visual dashboard, but it now includes developer access to the major data-platform capabilities through CredVault's own backend routes. Developers call CredVault methods; CredVault handles authentication, tenancy, audit logging, and the connection to orchestration, lineage, metadata, notebooks, agent sessions, realtime streams, dashboards, alerts, and pipelines.
Main Dashboard Areas
These are the main areas users see in the CredVault app sidebar.
| Sidebar area | What users do there | SDK status |
|---|---|---|
| Dashboard | View account, usage, clusters, keys, billing prompts, and system summaries | Partly covered through metrics, logs, clusters, and billing APIs; not a single SDK dashboard object |
| Notebook | Work with notebooks and code execution through the platform UI | Covered through notebooks |
| Agent Sessions | Use connected AI agent sessions from the platform | Covered through agentSessions / agent_sessions for authenticated session/event calls |
| ML Experiments | Work with experiment tracking and model workflows | CIE is covered; advanced experiment tracking remains dashboard-first |
| System Monitoring | View metrics, usage, performance, alerts, and dashboards | Covered through metrics, dashboards, alerts, logs, and realtime stream helpers |
| Orchestration | Explore orchestration and pipeline APIs | Covered through orchestration and pipelines |
| Lineage | Explore upstream and downstream data movement | Covered through lineage.getLineage / get_lineage and metadata lineage methods |
| Database Clusters | Create, inspect, and manage data clusters and collections | Partly covered through data, schema, backups, functions, and triggers |
| Metadata | Search and inspect platform metadata | Covered through metadata |
| API Keys | Create and manage API keys | Covered by apiKeys / api_keys with signed-in user token |
| Activity Logs | View and export user and tenant activity | Covered by logs |
| Migrations | Import, export, and test external data connections | Not covered by current SDKs |
| Billing & Plans | Manage plan choice, invoices, credits, limits, and payment methods | Not covered by current SDKs |
| Team Management | Manage team access and workspace membership | Dashboard/API first; not included in this SDK layer yet |
Current SDK Resource Map
The Node.js, Python, Go, and Java SDKs follow the same resource layout.
| SDK resource | What it does | Best credential |
|---|---|---|
auth | Sign up, sign in, and start user-authenticated sessions | Email/password or user token |
data | Query, insert, update, and delete collection documents; list clusters | API key for /api/v1/data; user token for cluster list |
cie | Upload datasets, list datasets, train models, list models, run predictions | User token |
webhooks | Create, update, delete, test, and inspect webhook deliveries | User token |
functions | Create, update, delete, test, execute, and read function logs | User token |
triggers | Manage cluster and collection trigger automation | User token |
backups | Create, restore, delete, and inspect backup state | User token |
schema | Analyze cluster schema and collection indexes | User token |
apiKeys / api_keys | List, create, revoke, and delete API keys | User token with owner/admin workspace role |
metrics | Read realtime, performance, security, and active collection metrics | User token |
logs | Read, filter, export, and summarize activity logs | User token |
notifications | List, mark read, mark all read, and delete notifications | User token |
settings | Intended for account settings; current backend coverage is limited | User token |
robots | Intended for robot/device APIs; current backend route shape differs | User token |
orchestration | List jobs, runs, assets, sensors, schedules, and launch orchestration jobs through CredVault | User token |
lineage | List namespaces, datasets, jobs, runs, versions, search metadata, and inspect lineage | User token |
metadata | List/search tables and databases and inspect metadata lineage | User token |
notebooks | Create, update, execute, export, checkpoint, undo, and redo CredVault notebooks | User token |
pipelines | Create, update, run, and inspect pipeline execution history | User token |
dashboards | Create dashboards, add visualizations, and fetch dashboard data | User token |
alerts | Create alerts, record metrics, and read alert history | User token |
realtime | List/stop realtime streams and build WebSocket stream URLs | User token |
agentSessions / agent_sessions | Create agent sessions and send session events through CredVault auth | User token |
Platform Integration Examples
Node.js:
const jobs = await credvault.orchestration.listJobs();
const run = await credvault.orchestration.runJob("daily_customer_refresh");
const lineage = await credvault.lineage.getLineage({
nodeId: "dataset:customers",
depth: 3,
direction: "DOWNSTREAM",
});
const tables = await credvault.metadata.search({ q: "customers" });
const execution = await credvault.pipelines.run("pipeline-id");
const dashboardData = await credvault.dashboards.data("dashboard-id");
Python:
jobs = client.orchestration.list_jobs()
run = client.orchestration.run_job("daily_customer_refresh")
lineage = client.lineage.get_lineage(
node_id="dataset:customers",
depth=3,
direction="DOWNSTREAM",
)
tables = client.metadata.search("customers")
execution = client.pipelines.run("pipeline-id")
dashboard_data = client.dashboards.data("dashboard-id")
Java:
String jobs = client.platform.orchestration.listJobs();
String run = client.platform.orchestration.runJob("daily_customer_refresh");
String tables = client.platform.metadata.search("customers", 10);
Go:
jobs, err := cv.Orchestration.ListJobs()
run, err := cv.Orchestration.RunJob("daily_customer_refresh")
tables, err := cv.Metadata.Search("customers", 10)
Production Base URL
The SDK source defaults to a local development server:
http://localhost:5000/api
Production applications must pass the deployed CredVault backend API URL:
https://<your-credvault-backend>/api
If a production app does not set the base URL, it will try to call the user's own computer and the request will fail.
API Keys And User Tokens
CredVault uses two different credential types.
Use an API key when your backend service needs direct data access:
curl https://<your-credvault-backend>/api/v1/data/customers \
-H "X-API-Key: <your-api-key>"
Use a user token when the action belongs to a signed-in dashboard user, such as managing clusters, functions, webhooks, API keys, logs, settings, or billing-related account state.
What Is Ready For SDK Use
These areas are the best fit for SDK-based development today:
- application data access through
/api/v1/data - cluster discovery through authenticated user sessions
- CIE datasets, model training, model listing, and predictions
- webhooks
- serverless functions
- triggers
- backups
- schema and indexes
- API key lifecycle operations
- metrics
- activity logs
- basic notifications
- orchestration jobs, runs, assets, sensors, and schedules
- lineage namespaces, datasets, jobs, runs, versions, search, and lineage graph queries
- metadata tables, databases, search, and metadata lineage
- CredVault notebook operations
- pipelines, dashboards, alerts, realtime stream management, and agent session/event calls
These are the features a developer building on CredVault should start with.
What Should Stay Dashboard Or REST API First
These areas are platform features, but they should not be presented as fully covered by the current SDKs yet:
- team and workspace member administration
- migrations and external connection testing
- billing, invoices, credits, and payment methods
- Coder Cloud IDE
- Pragma IDE
- advanced experiment tracking beyond the core SDK helpers
Users can still use these features in the dashboard or through documented REST APIs where available.
Recommended Wording For Users
Use this wording when describing the SDKs:
CredVault SDKs let developers build applications and automation on top of CredVault data, CIE, functions, webhooks, triggers, backups, schema, metrics, logs, API keys, notebooks, pipelines, dashboards, alerts, orchestration, lineage, metadata, realtime streams, and agent sessions.
Avoid saying:
The SDKs control every CredVault product and dashboard feature.
That would be inaccurate because some product areas are dashboard-first or integration-first today.
Choosing The Right Tool
| Goal | Recommended tool |
|---|---|
| Build an app that reads and writes CredVault data | SDK or REST API |
| Connect a backend service to CredVault | SDK with API key |
| Run data and AI workflows from a terminal | CIE CLI |
| Manage workspace users and billing | Dashboard |
| Explore notebooks, metadata, lineage, or orchestration | Dashboard integrations or SDK |
| Build a custom integration not covered by the SDK | REST API |
| Install and use Pragma or Coder | Their dedicated product pages and documentation |
The SDK should be presented as a developer acceleration layer for the stable core platform APIs. The dashboard remains the main place for broad platform management.