agentleFS
cognitedatacognitedata/builder-skills/skills/graph-viewer/SKILL.md

graph-viewer

Skill6 starsChanged 8 days agoApache-2.0View on GitHub

Integrate the reusable CDF graph viewer (useGraphViewer) into a Flows app by copying the local code bundle. Use when embedding a graph visualization, adding a knowledge graph, or showing CDF data model relationships and instances.

  • Installs packages

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 3 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not use this skill for static diagrams, pure dataflow visualizations, or non-CDF graphs.

    line 12
  2. Distinctive:

    Follow these steps in order.

    line 22
  3. Very specific:

    Inspect the target app: Read package.json and look at the existing folder structure (e.g.

    line 24
  4. Distinctive:

    If the repo already has a different feature/components layout or alias, mirror it.

    line 32
  5. Very specific:

    Only declare coreReverseQueries for relations the app must surface.

    line 84

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Graph Viewer

Use This When

The user wants to embed an interactive graph of a CDF data model — nodes, direct relations, edges, and reverse relations — inside a Flows app.

Do not use this skill for static diagrams, pure dataflow visualizations, or non-CDF graphs.

Prerequisites

  • The app is wrapped in @cognite/dune's <DuneProvider> so useDune() returns an authenticated SDK.
  • The target data model exists in CDF and you know its space, externalId, and version.
  • The app uses React 18+ and TypeScript.

Integration Workflow

Follow these steps in order. Adapt to the target repo's conventions instead of inventing new ones.

  1. Inspect the target app. Read package.json and look at the existing folder structure (e.g. src/features/*, src/components/*, path aliases like @/*).
  2. Install missing dependencies with the app's package manager (npm, pnpm, yarn, …). See the Dependencies table below for purposes and suggested versions. Reuse the React version already pinned by the app rather than upgrading it, and prefer any versions the repo already pins over the suggestions here.
  3. Copy the bundle into the app. Copy every file from skills/graph-viewer/code/ into an app-local feature folder, for example:

    src/features/graph-viewer/

    If the repo already has a different feature/components layout or alias, mirror it.

  4. Import from the local folder, never from @skills/.... With a typical @/* alias:

    import { useGraphViewer } from "@/features/graph-viewer";
  5. Render GraphCanvas inside a container with explicit dimensions (height is required — see the minimal example below).
  6. Run typecheck and build (tsc --noEmit, npm run build, etc.) and fix any path or type issues introduced by the copy.

Minimal Example

import { useGraphViewer } from "@/features/graph-viewer";

export function GraphPanel() {
  const { GraphCanvas, isLoading, error } = useGraphViewer({
    dataModel: { space: "my-space", externalId: "my-data-model", version: "1" },
    instance: { space: "my-instance-space", externalId: "pump-001" },
  });

  if (isLoading) return <div>Loading graph…</div>;
  if (error) return <div>Error: {error}</div>;

  return <GraphCanvas className="h-[600px] w-full" />;
}

Dependencies

Suggested versions reflect the latest published majors at the time of writing. They are starting points — if the target app already pins different versions, defer to the app.

PackageSuggested versionPurpose
react^18.2.0UI framework (peer; reuse the app's version)
@cognite/sdk^10.10.0CDF API client (instances, data models)
@cognite/dune^2.1.0Provides the authenticated SDK via useDune()
reagraph^4.30.8WebGL graph rendering engine
lucide-react^1.14.0Icon set used by the node-type legend

Example install (npm; adapt to the app's package manager):

npm install @cognite/sdk@^10.10.0 @cognite/dune@^2.1.0 reagraph@^4.30.8 lucide-react@^1.14.0

CDF Cost & Performance

Graph expansion can issue many CDF requests, especially with reverse relations. For large or unfamiliar data models, be conservative:

  • Set whitelistedRelationProps to the few properties the app actually needs to traverse.
  • Lower initialConnectionLimit (it is a hard maximum of connections fetched per expansion).
  • Lower maxNodes to bound the in-memory LRU buffer.
  • Only declare coreReverseQueries for relations the app must surface; each entry adds an extra query per expansion.

Tuples in coreReverseQueries are version-aware: [space, viewExternalId, viewVersion, propertyName, isList].

Advanced Reference

For full configuration tables, return-value docs, layouts, theming, and richer examples, read code/README.md.

For implementation details, inspect the source files under code/.

Verification Checklist

  • The app is wrapped in <DuneProvider>.
  • All files from skills/graph-viewer/code/ were copied into an app-local folder.
  • Imports point to the app-local folder (e.g. @/features/graph-viewer), not @skills/....
  • @cognite/dune, @cognite/sdk, reagraph, and lucide-react are present in package.json.
  • The container that renders <GraphCanvas> has an explicit height.
  • tsc --noEmit and the app's build both pass.
  • No references to dune-industrial-components were introduced.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.