Content
75%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a well-organized overview/index skill that correctly assumes Claude's competence and pushes detail out to the npm package's shipped docs, guides, and templates, with concrete commands and paths throughout. Its main weaknesses are the absence of a minimal executable bot example, an unordered presentation of where to look first, and minor marketing verbosity in the tagline and guide descriptions.
Suggestions
Add a minimal executable bot snippet (or a concrete pointer to a quick-start example inside node_modules/chat/docs/) to the scaffold section so the first action an agent takes is copy-paste ready.
Present an explicit ordered flow — e.g., '1. Verify chat is installed (else npm i chat) 2. Pick an adapter via @chat-adapter/catalog 3. Follow the matching guide below' — instead of parallel sections, so the agent knows where to look first.
Trim the marketing tagline ('Write your bot logic once, deploy everywhere') and shorten each guide entry to its title plus the platform/SDK stack to reclaim tokens without losing routing value.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is efficient and assumes Claude's competence — it never explains what a bot or npm package is — but includes minor trimmable tokens: the marketing tagline 'Write your bot logic once, deploy everywhere', and multi-clause guide descriptions (e.g., 'Covers project setup, tool definitions, streaming responses, deployment to Vercel, and scaling tool selection with toolpick') that could be tightened to title plus key stack. This fits anchor 4 (efficient with minor instances that could be trimmed) rather than anchor 5, where every token would earn its place. | 4 / 5 |
Actionability | Concrete, executable guidance is present throughout: the commands 'npx create-chat-sdk --help' and 'npm i chat', specific paths ('node_modules/chat/docs/', 'node_modules/chat/resources/templates.json'), named catalog imports ('listAdapters', 'getCatalogEntry', 'getFeatureSupport'), and per-guide file paths. It falls short of anchor 5 only because there is no copy-paste-ready minimal bot example; it exceeds anchor 3 because what is given is executable and specific rather than high-level hints or pseudocode. | 4 / 5 |
Workflow Clarity | The flow is discernible (scaffold → verify the package is installed → consult docs/guides/templates → select an adapter via the catalog), and it includes an explicit conditional checkpoint: 'If those paths do not exist, the `chat` package is not installed in the project yet.' No destructive or batch operations are involved, so the validation cap does not apply. It matches anchor 4 rather than 5 because sections are presented as a parallel catalog rather than an ordered decision flow, and there is no guidance on when to prefer llms.txt vs. the catalog vs. the bundled docs. | 4 / 5 |
Progressive Disclosure | The skill ships no bundle files and delegates all detail to clearly signaled, one-level-deep external sources — 'node_modules/chat/docs/', the [Chat SDK llms.txt file](https://chat-sdk.dev/llms.txt), templates.json, and the website/KB links — with well-organized sections. It sits at anchor 4 rather than 5 because the ~30-line auto-generated resources catalog (between the RESOURCES:START/END markers) is inlined in SKILL.md, content that arguably belongs in a separate reference file even though it serves as the skill's discovery index. | 4 / 5 |
Total | 16 / 20 Passed |