Content
80%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 an efficient, well-organized reference of executable templates for building hooks, with a clear implicit workflow and a manual testing step. Its main weaknesses are the placeholder-laden TypeScript handler template and the absence of error-recovery guidance when a test fails or a hook's output is malformed.
Suggestions
Make the TypeScript template fully executable by defining the readStdin() helper (e.g. a small snippet that reads stdin to a string) so the handler pattern is copy-paste ready.
Add a short error-recovery note to the Testing section: what to check when a hook produces no output or invalid JSON (exit codes, stderr, npx tsx failures), forming a validate-then-fix feedback loop.
Include one complete worked example (e.g. a filled-in PreToolUse handler with a real HookInput shape) so the event-specific field structure is concrete rather than implied.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and template-driven — shell/TypeScript/JSON examples plus a terse event list — with no explanation of concepts Claude already knows and no padding. Re-checked against anchor 4 (minor over-explanation to trim): none found, so anchor 5 (every token earns its place) is the best fit. | 5 / 5 |
Actionability | The shell wrapper ("cat | npx tsx <handler>.ts"), the test command, and the registration JSON are concrete and near copy-paste ready, but the TypeScript handler is a skeleton with placeholders (undefined readStdin(), "// Process input", "EventName"). That places it at anchor 4 (mostly executable, minor gaps) rather than anchor 5 (fully copy-paste ready) or anchor 3 (pseudocode-level guidance). | 4 / 5 |
Workflow Clarity | The skill lays out a coherent sequence — write the shell wrapper, implement the TypeScript handler, register in settings.json, then test with the echo command — and includes an explicit verification step ("Test hooks manually"). It is not a destructive/batch operation so the validation cap does not apply; it falls short of anchor 5 only because there is no error-recovery feedback loop (e.g. what to do when the test fails or a hook's output is malformed). | 4 / 5 |
Progressive Disclosure | This is a compact (~65-line), single-file skill with no bundle files (references/, scripts/, and assets/ do not exist), and its sections are clearly headed (Pattern, Templates, Events, Testing, Registration) with nothing that needs splitting out. Per the simple-skill guidance this matches anchor 5 (well-organized sections, no external references needed); anchor 4 would require organization gaps, and none are present. | 5 / 5 |
Total | 18 / 20 Passed |