agents-sdk

Build AI agents on Cloudflare Workers using the Agents SDK. Load when creating stateful agents, durable workflows, real-time WebSocket apps, scheduled tasks, MCP servers, or chat applications. Covers Agent class, state management, callable RPC, Workflows integration, and React hooks. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.

2.20x

Quality

93%

Does it follow best practices?

Impact

99%

2.20x

Average score across 3 eval scenarios

Securityby

Low

Low-risk findings worth noting

The canonical home for this skill is agents-sdk in cloudflare/skills

Quality

Content

87%Weight 40%Scale 1-3

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is an efficient, actionable overview with executable examples and well-structured one-level-deep references. Its only gap is workflow clarity: the build flow lacks explicit validation checkpoints beyond the initial install check.

Suggestions

Add an explicit validation/verification checkpoint after the Wrangler config and Agent class steps (e.g., `npx wrangler dev` smoke test or `wrangler deploy --dry-run`) to close the workflow-clarity gap.

Remove the redundancy between the "Capabilities" bullet list and the Core APIs / Retrieval Sources tables, or collapse them into a single table to tighten token use.

Add a brief ordered "Build an agent" workflow (install → configure wrangler → define Agent class → route → connect client) so the sequenced steps are explicit rather than implied by section order.

Dimension	Reasoning	Score
Conciseness	Lean reference style—tables, config, and complete code with no explanation of concepts Claude already knows. The "Capabilities" bullets overlap slightly with the Core APIs and Retrieval Sources tables, but every token still earns its place.	3 / 3
Actionability	Provides fully executable, copy-paste-ready TypeScript (Agent class), TSX (React client), JSONC (wrangler config), and bash install commands, plus a specific Core APIs table. Not score 2 because examples are complete rather than pseudocode.	3 / 3
Workflow Clarity	A sequence is implied by section order plus the explicit "FIRST: Verify Installation" checkpoint and retrieval-first directive, but there are no validation feedback loops for the core agent-building task. Not score 3 because checkpoints after writing/configuring code are missing or only implicit.	2 / 3
Progressive Disclosure	SKILL.md is a concise overview with a References section linking to seven real, one-level-deep reference files (verified present in references/), each clearly signaled with a short description. Not score 2 because content is appropriately split with easy navigation rather than inlined.	3 / 3
	Total	11 / 12 Passed

Description

100%Weight 40%Scale 1-3

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific, trigger-rich, and explicitly answers both what the skill does and when to load it, all scoped to a distinct Cloudflare niche. It cleanly matches the strongest anchor on every dimension.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete capabilities—"stateful agents, durable workflows, real-time WebSocket apps, scheduled tasks, MCP servers, or chat applications" and "Agent class, state management, callable RPC, Workflows integration, and React hooks"—matching the anchor for listing several specific concrete actions.	3 / 3
Completeness	Explicitly answers both: what ("Build AI agents on Cloudflare Workers using the Agents SDK") and when ("Load when creating stateful agents, durable workflows...") with an explicit "Load when" trigger clause.	3 / 3
Trigger Term Quality	Natural user-facing terms ("AI agents", "Cloudflare Workers", "WebSocket apps", "MCP servers", "scheduled tasks") with good coverage; not merely technical jargon. Not score 2 because common variations are present rather than missing.	3 / 3
Distinctiveness Conflict Risk	Clear niche scoped to Cloudflare Workers + Agents SDK with distinct triggers, making conflict with other skills unlikely. Not score 2 because the scope is tightly bound to a specific platform/SDK rather than generic.	3 / 3
	Total	12 / 12 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: openai/plugins
Path: plugins/cloudflare/skills/agents-sdk/SKILL.md
Commit: 11c74d6

Reviewed: 17 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.