Content
92%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 thin-router skill body: unambiguous prioritized routing rules, concrete artifact heuristics for the 'auto' case, and strict non-duplication of destination logic. The only notable flaw is mild redundancy between the subcommand table, routing rule 1, and the UX Convention section.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes Claude's competence, but rule 1 ('If the user explicitly says claude or codex, route directly') duplicates the Subcommands table, and the UX Convention section repeats the intro's document-vs-history distinction — minor trimmable redundancy that keeps it below the 'every token earns its place' anchor. | 4 / 5 |
Actionability | For an instruction-only router, the guidance is fully executable: a dispatch table with exact destination skill names, concrete artifact patterns ('~/.claude or Claude memory/session JSONL artifacts'), a copy-paste clarification question, and an explicit fallback for unsupported sources covering all common cases. | 5 / 5 |
Workflow Clarity | The prioritized decision rules (explicit target → path-based inference → ask one clarification → state limitations) plus the Execution Contract make the single routing task unambiguous; no destructive or batch operations exist, so the validation cap does not apply and the simple-skill exception holds. | 5 / 5 |
Progressive Disclosure | Under 50 lines with no external bundle files needed; delegation to destination skills (claude-history-ingest, codex-history-ingest, wiki-ingest) is one level deep and clearly signaled, and the sections are well-organized, satisfying the simple-skill standard for a full score. | 5 / 5 |
Total | 19 / 20 Passed |