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, chat applications, voice agents, or browser automation. Covers Agent class, state management, callable RPC, Workflows, durable execution, queues, retries, observability, and React hooks. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
89
86%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Advisory
Suggest reviewing before use
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is an excellent skill description that clearly defines its scope (Cloudflare Workers Agents SDK), lists comprehensive concrete capabilities, and provides explicit trigger conditions via the 'Load when' clause. It uses proper third-person voice throughout and includes a helpful note about preferring Cloudflare docs over pre-trained knowledge. The description is information-dense without being padded or vague.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions and concepts: 'Build AI agents', 'stateful agents', 'durable workflows', 'real-time WebSocket apps', 'scheduled tasks', 'MCP servers', 'chat applications', 'voice agents', 'browser automation', plus technical specifics like 'Agent class', 'callable RPC', 'Workflows', 'queues', 'retries', 'observability', and 'React hooks'. | 3 / 3 |
Completeness | Clearly answers both 'what' (build AI agents on Cloudflare Workers using the Agents SDK, covering Agent class, state management, RPC, Workflows, etc.) and 'when' ('Load when creating stateful agents, durable workflows, real-time WebSocket apps, scheduled tasks, MCP servers, chat applications, voice agents, or browser automation'). The 'Load when' clause serves as an explicit trigger guidance. | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural terms a user would say: 'AI agents', 'Cloudflare Workers', 'Agents SDK', 'WebSocket', 'scheduled tasks', 'MCP servers', 'chat applications', 'voice agents', 'browser automation', 'state management', 'durable execution'. These are terms developers would naturally use when seeking help with this technology. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive due to the specific combination of 'Cloudflare Workers' and 'Agents SDK'. The niche is clearly defined and unlikely to conflict with generic coding skills, general AI agent skills, or other cloud platform skills. The mention of Cloudflare-specific concepts like 'durable execution' and 'Cloudflare docs' further narrows the scope. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
72%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured skill that excels at progressive disclosure and actionability, providing executable code examples and clear reference organization. The main weaknesses are moderate verbosity (the capabilities list largely duplicates the retrieval table) and implicit rather than explicit workflow sequencing with no validation checkpoints after setup steps.
Suggestions
Remove or significantly condense the '## Capabilities' bullet list since the retrieval sources table already maps topics to their docs — this would save ~20 lines of redundant content.
Add an explicit 'Verify it works' step after the wrangler configuration and agent class sections (e.g., 'Run `npx wrangler dev` and visit `/agents/counter/test` to confirm the agent responds').
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The large retrieval sources table and capabilities list are somewhat redundant with each other and with the reference links at the bottom. The core content (Agent class example, APIs table, React client) is efficient, but the doc could be tightened by removing the full capabilities bullet list since the retrieval table already covers the same topics. | 2 / 3 |
Actionability | Provides fully executable code examples (Agent class, React client, wrangler config, install commands), a concrete Core APIs table with exact method signatures, and specific gotchas. The code is copy-paste ready and covers the most common patterns. | 3 / 3 |
Workflow Clarity | The skill provides a clear setup sequence (verify installation → configure wrangler → create agent class → set up routing → connect client), but lacks explicit validation checkpoints. There's no 'verify it works' step after deployment, no error recovery guidance for common failures, and the workflow is implicit rather than explicitly sequenced. | 2 / 3 |
Progressive Disclosure | Excellent progressive disclosure with a concise overview containing essential code examples, followed by well-organized one-level-deep references split by domain (Core, Chat & Streaming, Background Processing, Integrations, Experimental). The retrieval sources table also provides clear navigation to external docs. | 3 / 3 |
Total | 10 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Repository
- cloudflare/skills
- Commit
7c449de
- Reviewed
Table of Contents
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.