---
title: "Cloudflare Agents SDK: review, setup & limitations · Undominated.ai"
canonical: https://undominated.ai/skills/cloudflare-agents-sdk/
description: "Guides stateful Cloudflare agent applications through current SDK documentation, persistence, scheduling and approval-aware tool flows."
---

# Cloudflare Agents SDK: review, setup & limitations · Undominated.ai

> Guides stateful Cloudflare agent applications through current SDK documentation, persistence, scheduling and approval-aware tool flows.

[← Explore all skills](/skills/)

AGENT DEVELOPMENT / Cloudflare

# Cloudflare Agents SDK

Guides stateful Cloudflare agent applications through current SDK documentation, persistence, scheduling and approval-aware tool flows.

 [See setup guidance ↓](#setup)[Original source ↗](https://github.com/cloudflare/skills/tree/53395984c3ea1925d00fce03b266112302f050e6/skills/agents-sdk)

SOURCE REVIEW

 Reviewed 2026-10-15
 Evidence 8 linked sources
 Publisher Cloudflare
 Licence [Apache-2.0 ↗](https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/LICENSE)
 Revision 53395984c3ea
 [Read what was—and wasn’t—checked ↓](#review)

“Build, debug, or review Cloudflare Agents SDK applications using the agents package.”

 [Cloudflare · upstream description ↗](https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/skills/agents-sdk/SKILL.md) Our analysis follows below.

01 / THE REASONING

## Why this made the selection.

 - Treats installed SDK versions and current docs as the basis for implementation.
- Explains the difference between synchronized state, database history, chat-tool approvals and durable-workflow pauses.

### A good fit for

 - Building a persistent agent on Cloudflare Workers
- Adding scheduled work or an approval step to an agent application

### Weigh up before choosing

 - This is specific to Cloudflare’s Agents SDK and runtime, not a general multi-provider agent framework.
- Experimental browser, voice and higher-level APIs need current documentation and may change.
- Installing the skill does not configure authentication, deploy resources or verify an agent application.

02 / THE REVIEW RECORD

## What we actually inspected.

Source review has boundaries. A clear record is more useful than a “safe” badge.

### Material inspected

 - skills/agents-sdk/SKILL.md
- skills/agents-sdk/references/human-in-the-loop.md
- skills/agents-sdk/references/state-scheduling.md
- LICENSE
- skills/agents-sdk/SKILL.md and both inspected references compared with the previous review revision b052c32b: every changed line is a documentation link, now to an index.md Markdown page; the instructions are otherwise unchanged.

### Our findings

 - The definition checks SDK installation, bindings and migrations and warns against editing old migration tags.
- The state reference separates validation of proposed updates from reactions to accepted changes.
- The human-in-the-loop reference distinguishes approval responses from client tool outputs and calls for timeout handling before gated work.
- The definition opens with "Prefer retrieval over pre-training" and links each topic, from state and scheduling to MCP and experimental voice, to a Markdown (index.md) Cloudflare docs page.

### Not established by this review

 - Source inspection only; no skill installation or workflow execution was performed.
- Not executed: it was not installed or run.

The review applies to the material and revision named here. A newer upstream release can change its behavior.

03 / PUT IT TO WORK

## Add a skill to your workflow.

[Upstream setup instructions ↗](https://github.com/vercel-labs/skills/blob/13e4063a1cf913f5606d57d42ab83a86f5001e04/README.md)

DOCUMENTED COMMAND

 npx skills add cloudflare/skills/skills/agents-sdk Copy command ↗

Run in your terminal to choose your AI tools, project or global scope, and installation method. This short command follows the current upstream version; the review covers the revision shown above.

 Optional: install the reviewed revision

Keeps the original source and installer version where recorded, while leaving your AI tools and installation scope for you to choose.

 npx skills add https://github.com/cloudflare/skills/tree/53395984c3ea1925d00fce03b266112302f050e6/skills/agents-sdk Copy reviewed-revision command ↗ - Install the complete skill directory with its references and helpers; a copied SKILL.md alone may be incomplete.
- Choose your AI tools and project or global scope in the installer, then review its permissions before using the skill.

### Before you start

 - Node.js/npm for this installer
- An agent that supports the Agent Skills format
- A configured Workers project
- Cloudflare account access for deployment and provider credentials where used

### Compatibility

Cloudflare Workers and agents package · Compatible JavaScript/TypeScript coding agents

### Read and edit agent code, bindings and migrations

 - Read and edit agent code, bindings and migrations
- Optional deployment to persistent cloud resources
- Configured agents may store data, invoke tools or contact external model services

### Cost model

The skill is Apache-licensed; Workers resources, storage and any model/provider services are billed independently.

04 / FOLLOW THE EVIDENCE

## The source trail.

Our notes are separate from the original resource. Check upstream before adopting a new version.

 - [Skill definition at reviewed revision ↗](https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/skills/agents-sdk/SKILL.md) Checked 2026-10-13 https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/skills/agents-sdk/SKILL.md Supports: summary, upstreamDescription, whySelected, bestFor, limitations, review, compatibility, access
- [Reviewed reference or helper: references/human-in-the-loop.md ↗](https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/skills/agents-sdk/references/human-in-the-loop.md) Checked 2026-10-13 https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/skills/agents-sdk/references/human-in-the-loop.md Supports: review, limitations, access
- [Reviewed reference or helper: references/state-scheduling.md ↗](https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/skills/agents-sdk/references/state-scheduling.md) Checked 2026-10-13 https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/skills/agents-sdk/references/state-scheduling.md Supports: review, limitations, access
- [Upstream license or licensing declaration ↗](https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/LICENSE) Checked 2026-10-13 https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/LICENSE Supports: license
- [Publisher installation and project documentation ↗](https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/README.md) Checked 2026-10-13 https://github.com/cloudflare/skills/blob/53395984c3ea1925d00fce03b266112302f050e6/README.md Supports: install, compatibility, access
- [Skills installer source formats and selective installation ↗](https://github.com/vercel-labs/skills/blob/13e4063a1cf913f5606d57d42ab83a86f5001e04/README.md) Checked 2026-10-13 https://github.com/vercel-labs/skills/blob/13e4063a1cf913f5606d57d42ab83a86f5001e04/README.md Supports: install
- [Installer implementation: full commit SHA fallback ↗](https://github.com/vercel-labs/skills/blob/13e4063a1cf913f5606d57d42ab83a86f5001e04/src/git.ts) Checked 2026-10-13 https://github.com/vercel-labs/skills/blob/13e4063a1cf913f5606d57d42ab83a86f5001e04/src/git.ts Supports: install
- [2026-10-15 W34 source re-review at the existing pin ↗](https://raw.githubusercontent.com/cloudflare/skills/53395984c3ea1925d00fce03b266112302f050e6/skills/agents-sdk/SKILL.md) Checked 2026-10-15 https://raw.githubusercontent.com/cloudflare/skills/53395984c3ea1925d00fce03b266112302f050e6/skills/agents-sdk/SKILL.md Supports: upstreamDescription, review.notTested, review.reviewedAt

KEEP COMPARING

## Other approaches to consider.

Related by category or shared topics. These are alternatives to inspect, not a measured quality order.

 [### Agent Development for Claude Code Plugins ↗ Explains how to write a Claude Code plugin agent file (frontmatter, trigger description, system prompt) and ships a Bash script that checks one agent file.](/skills/anthropic-agent-development/)[### Context engineering ↗ Orders agent context as a hierarchy instead of dumping the repository into the prompt.](/skills/addyosmani-context-engineering/)[### Docker Agent run ↗ Pins Docker's skill for docker agent run, including the safety modes and the exception that a local stdio MCP server stays outside the sandbox VM.](/skills/docker-agent-run/)

 [AI Tools ↗](/tools/)[Skills ↗](/skills/)[Agents ↗](/agents/)[MCP Servers ↗](/mcp-servers/)[Workflows ↗](/workflows/)

## Continue your investigation

 - [Choose a toolkit](/tools/)
- [Inspect execution hosts](/agents/)
- [Check tool access](/mcp-servers/)
- [Check project dependencies](/check-repo/)
