agentleFS
Sign inStart free
geoai-latamgeoai-latam/SKILL_JS5_ARCGIS/skills/arcgis-ai-components/AGENTS.md

SKILL_JS5_ARCGIS / arcgis-ai-components

AGENTS.md1 starsChanged 6 months agoMITView on GitHub

Quick-reference for AI Components. Read before adding an assistant. SDK 5.1. 1. Chat assistant on a 2D web map? → arcgis-assistant + built-in agents. 2. Capabilities? → navigation / data-exploration / help / knowledge (5.1+). (No layer-filter — filtering is in data-exploration.) 3. Domain logic, custom tools, or a 3D scene? → custom agent: agent-utils (FunctionAgent, LLMAgent, workflows — 5.1) or a hand-written LangGraph graph. - [ ] Pin @arcgis/ai-components exactly (beta; 5.1 broke the orchestrator↔agent contract). - [ ]…

The file

We split this AGENTS.md into the rules it gives an agent

1 rule. About 1 is common, 0 are distinctive, and 0 are very specific to this project.

  • ≈1

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Common:

    No API keys, no Enterprise yet.

    line 21

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md 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
AGENTS.md

Agent Guide: arcgis-ai-components (BETA)

Quick-reference for AI Components. Read before adding an assistant. SDK 5.1.

Quick Decision Tree

  1. Chat assistant on a 2D web map? → arcgis-assistant + built-in agents.
  2. Capabilities? → navigation / data-exploration / help / knowledge (5.1+). (No layer-filter — filtering is in data-exploration.)
  3. Domain logic, custom tools, or a 3D scene? → custom agent: agent-utils (FunctionAgent, LLMAgent, workflows — 5.1) or a hand-written LangGraph graph.

Setup Checklist

  • Pin @arcgis/ai-components exactly (beta; 5.1 broke the orchestrator↔agent contract).
  • Import from @arcgis/ai-components/components/<tag> — dist/components/... is not exported (Vite fails).
  • Built-in agents need a web map (item-id): otherwise WebMap portal item is missing.
  • With pnpm, add @langchain/core@^1.3.0 directly (else Missing "./utils/gateway" specifier).
  • Register ≥1 agent (built-in or custom) — the assistant needs one to work.
  • Bind built-in agents via reference-element → arcgis-map only (2D, no SceneView).
  • Auth: named-user OAuth sign-in to an ArcGIS Online org with AI assistants enabled. No API keys, no Enterprise yet.
  • data-exploration needs AI vector embeddings generated on the web map's layers.
  • Custom agents: bundler + @langchain/langgraph @langchain/core zod; createGraph returns an uncompiled graph; workspace exposes outputMessage (a 5.1 FunctionAgent/LLMAgent does both for you).
  • Scope/guard the data the assistant can access.

Anti-patterns (one-liner reminders)

  • ❌ Expecting arcgis-scene/SceneView to work with reference-element (2D only).
  • ❌ Floating beta versions (^5).
  • ❌ import "@arcgis/ai-components/dist/components/..." (type-checks, fails in Vite/Node).
  • ❌ Built-in agents on a map with only basemap= (no web map → orchestrator never initializes).
  • ❌ Assistant with no registered agent.
  • ❌ Assuming API keys / public accounts / Enterprise work for the assistant.
  • ❌ Custom agent whose workspace/nodes don't set outputMessage (blank chat).

3D scenes (workaround)

The assistant won't bind a scene, but a custom agent's tools are your own code. Register only custom agents (so reference-element isn't required), pass the SceneView in context, and drive it from tools. See SKILL Concept 4 + examples.md Example 4.

Files in This Skill

  • SKILL.md / SKILL.es.md — full guides
  • references/api-detail.md — assistant / agents / custom-agent API
  • references/examples.md — setup + custom-agent + 3D examples
  • references/pitfalls.md — extended troubleshooting
  • scripts/snippet-assistant.ts
  • (repo) CUSTOM_AGENTS_GUIDE.md — deep LangGraph build guide

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.

Check your setup, free