Content
93%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 a concise, actionable overview with copy-paste CLI examples and well-structured one-level-deep references. The only minor gap is that error-recovery feedback steps are delegated to the reference rather than surfaced inline.
Suggestions
Add a short inline note on the key exit codes / business codes (e.g. 100004 rate-limited, exit 3 network) and the retry guidance so the SKILL.md body carries a minimal validation checkpoint.
Consider a one-line 'response shape' cue (e.g. results under .data[], percentages pre-formatted) in the body so Claude can parse output without always opening references/cli.md.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean — tables for commands/chains and terse rule bullets — and assumes Claude's competence, avoiding explanations of what meme tokens or bonding curves are conceptually. | 5 / 5 |
Actionability | Provides copy-paste-ready executable CLI invocations with full JSON parameter examples, exact required arguments, and concrete enum mappings for rankType/sort/protocol. | 5 / 5 |
Workflow Clarity | The two-command sequence is unambiguous, but error/validation feedback steps (e.g. handling rate-limit code 100004, network exit 3) live only in references/cli.md rather than inline, leaving a minor checkpoint gap. | 4 / 5 |
Progressive Disclosure | SKILL.md is an overview that points to a single, clearly signaled one-level-deep references/cli.md (a real bundle file) for the full reference; sections are well organized and the bulk content is appropriately split. | 5 / 5 |
Total | 19 / 20 Passed |