Content
93%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.
An exemplary lean reference: every section is a set of executable commands mapped to real bundle scripts, with a concrete end-to-end example flow. The only meaningful gap is the absence of explicit verification checkpoints in the multi-step flows, though for read-only browsing operations this is a minor concern.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Nearly every line is a command, flag, path, or one-line note (e.g. "./scripts/start.js --profile # Copy your profile into isolated cache"), with no primer on CDP or browser concepts Claude already knows; the only prose (profile behavior notes, the string-escaping tip) delivers non-obvious operational facts, so every token earns its place. | 5 / 5 |
Actionability | All examples are fully executable, copy-paste-ready commands with realistic arguments (e.g. "./scripts/eval.js 'document.querySelectorAll("a").length'" and "BROWSER_BIN=/path/to/chrome ./scripts/start.js"), covering the common cases for each tool including flag variants and environment overrides. | 5 / 5 |
Workflow Clarity | The "Quick Mobile Debug Flow" section gives a clear ordered multi-step sequence (start, navigate, emulate, reload, dismiss cookies, screenshot) with an inline comment explaining the reload step, and one error-handling note ("start will fail instead of reusing" an occupied port), but there are no explicit verify/validate checkpoints (e.g. confirming the screenshot reflects the mobile viewport). Anchor 5 requires explicit validation steps and feedback loops; anchor 3 would need an incoherent or implicit sequence, and the destructive/batch cap does not apply since these are read-only browser interactions. | 4 / 5 |
Progressive Disclosure | The body is organized as one tight section per bundle script, every referenced script exists in ./scripts/, and bulk detail is appropriately deferred to runtime commands ("./scripts/emulate.js --list") instead of being inlined; nothing that belongs in a separate file is in the SKILL.md, so there are no organization gaps. | 5 / 5 |
Total | 19 / 20 Passed |