Content
78%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.
A tight, highly actionable skill body that stays lean and gives exact commands and markup, with a clear step sequence and sensible contingencies. Its main weaknesses are the absence of any post-edit validation step and that the two bundled files it depends on are not actually present in the skill directory.
Suggestions
Add an explicit validation checkpoint before reporting the path, e.g. confirm the embedded JSON blob parses (or spot-check that the hero total renders from overall.input_tokens.total) so a truncated or malformed embed is caught.
Ship the referenced bundle files (analyze-sessions.mjs and template.html) in the skill directory, or document their flags and template placeholders inline so the skill is self-contained and verifiable.
Replace the <skill-dir> placeholder guidance with a concrete resolution note (e.g. how to derive the skill directory at runtime) so the commands are truly copy-paste ready.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~40-line body is lean and imperative throughout — numbered steps, exact shell commands, exact HTML markup for findings, and a short Notes section — with zero explanation of concepts Claude already knows and no padding, matching the 'every token earns its place' anchor. | 5 / 5 |
Actionability | Concrete, near-executable guidance: exact commands ("node <skill-dir>/analyze-sessions.mjs --json --since 7d > /tmp/session-report.json", "cp <skill-dir>/template.html ./session-report-$(date +%Y%m%d-%H%M).html"), verbatim output markup, and named JSON fields to read. It falls short of fully copy-paste-ready because of the unresolved <skill-dir> placeholder and reliance on a bundled script whose flags cannot be verified (the referenced analyze-sessions.mjs is not present in the skill directory). | 4 / 5 |
Workflow Clarity | A clear 5-step sequence with a contingency (trim top_prompts/cache_breaks if JSON >2MB) and a preservation guard ("use Edit, not Write — preserve the template's JS/CSS"), but there is no explicit validation checkpoint (e.g. verifying the embedded JSON parses or the report renders) before reporting the file path. This is a minor validation gap rather than the feedback-loop-rich 5 anchor; the operations are non-destructive, so the destructive/batch cap at 3 does not apply. | 4 / 5 |
Progressive Disclosure | The body itself is well organized into Steps/Notes and appropriately delegates implementation to two bundled files, but those referenced files (analyze-sessions.mjs and template.html, referenced via "<skill-dir>/...") do not exist in the skill directory — no scripts/, references/, or assets/ bundles are present — so the references cannot be verified and the skill's structure is incomplete as shipped, fitting the 'some structure but could be better organized / references not verifiable' middle anchor. | 3 / 5 |
Total | 16 / 20 Passed |