Content
92%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
An actionable, concise, well-sequenced verification workflow with explicit retry and validation feedback loops and copy-paste-ready commands. The only mild gap is that the extended server-correlation diagnostics are inline rather than split into a one-level-deep reference file.
Suggestions
Move the 'Diagnosing a Hanging or Stuck Page' section and its correlation table into a referenced reference file (e.g. DIAGNOSING.md) so the core flow stays lean and the deep troubleshooting is one click away.
If the workflow/devkit log commands are environment-specific, flag that precondition inline so the steps remain actionable when those tools are absent.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and instruction-forward — short bash command blocks and a tight checklist — with no padding explaining what a dev server or browser console is; every section earns its place. | 3 / 3 |
Actionability | Provides fully executable `agent-browser` commands and concrete JS eval snippets for each check, with specific selectors and copy-paste-ready bash, e.g. `agent-browser eval 'document.body.innerText.trim().length > 0 ? "HAS_CONTENT" : "BLANK"'`. | 3 / 3 |
Workflow Clarity | A clear numbered verification checklist is followed by an explicit failure loop ('fix the issue in code', 'Re-run verification (max 2 retry cycles)') and a correlate-browser-with-server diagnostic table, giving explicit validation checkpoints and feedback loops. | 3 / 3 |
Progressive Disclosure | Content is well-organized into clearly signaled sections, but the diagnostic 'Correlate Browser + Server' table and server-log steps are fairly large inline blocks that, with no bundle files present, could be split into a referenced reference; still, it stays one level deep with no nested references. | 2 / 3 |
Total | 11 / 12 Passed |