Content
63%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.
Highly actionable and clearly sequenced, but the SKILL.md body is far too verbose for an overview, inlining reference-grade detail instead of deferring it to the bundle files that already exist. Tightening the body and pushing security/performance/extension depth into the references would materially raise the score.
Suggestions
Trim the body to a concise overview: remove the repeated ASCII connection/status output boxes, the 'What You Can Do' IntelliSense/file-explorer lists, the mobile usage tips, the success checklist, and the code-server-vs-ttyd comparison — these are either known concepts or belong in reference files.
Move the full Security Essentials (threat model, do/don't lists, advanced security), Performance Tips, and extension-compatibility detail into their existing reference guides and replace them with one-line pointers.
Replace the inline Templates section and the verbose 'Reference Guides'/'Automation Scripts' duplications with a single consolidated reference map; keep only the install→configure-auth→configure-resources→start→stop→status spine in SKILL.md.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | At ~1139 lines the body is noticeably verbose, padding concepts Claude already knows (explaining IntelliSense, file explorer, git integration) and repeating large ASCII output boxes, mobile tips, and checklists that earn few tokens for their length. | 2 / 5 |
Actionability | Fully executable guidance throughout: exact scripts to run, real extension IDs, copy-paste-ready config snippets, and concrete troubleshooting commands covering the common cases. | 5 / 5 |
Workflow Clarity | A clearly sequenced 8-step workflow with verification at most stages (health-check.sh, 'Verify new limits', 'Verify Stopped', status-after-stop) and per-step troubleshooting; a couple of destructive/stop checkpoints rely on script behavior rather than an explicit validate-then-proceed gate. | 4 / 5 |
Progressive Disclosure | References and scripts are real and well-signaled one level deep, but the body inlines large blocks that belong in those references (full security threat model, performance tips, extension compatibility, multiple full terminal output boxes, templates), leaving the overview far longer than a SKILL.md should be. | 3 / 5 |
Total | 14 / 20 Passed |