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.
The body is a strong example of a CLI-wrapper skill: fully executable commands with real example arguments that match the actual bundled script, a clearly sequenced procedure with error-recovery guidance and a verification step, and appropriate separation of usage documentation from the implementation script. The only notable weakness is token-level repetition of the invocation path and dual command listings (Quick Reference plus Procedure).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body assumes Claude's competence throughout — no padding about what Hyperliquid, funding, or order books are — and every section carries skill-specific information (commands, flags, endpoint quirks, pitfalls). It misses a 5 because the full `~/.hermes/skills/...` invocation path is repeated in every Procedure example and all 12 command syntaxes appear in both Quick Reference and Procedure. | 4 / 5 |
Actionability | Every command is fully executable and copy-paste ready with flags, defaults, and concrete example arguments (`candles BTC --interval 1h --hours 72`, `export BTC --end-time-ms 1760000000000`, `review --coin BTC --hours 168`), and all 12 subcommands and flags were verified to exist in `scripts/hyperliquid_client.py`. The Verification section adds a runnable check with expected output. | 5 / 5 |
Workflow Clarity | The Procedure is clearly sequenced (discover dexs/markets, pull history, inspect live book, review account, fills, trade review, export), step 6 chains an explicit deeper-analysis flow (review → fills/orders → candles/funding), and the Pitfalls section provides error-recovery feedback ("iterate with later `startTime` values" for rate-limited/capped windows). A final Verification step with expected output completes the loop; the skill is read-only, so the destructive-operation cap does not apply. | 5 / 5 |
Progressive Disclosure | The bundle is appropriately split: the 60KB implementation lives in `scripts/hyperliquid_client.py` (the referenced helper path matches the bundle file), while SKILL.md stays at overview/usage level with well-organized sections (When to Use, Prerequisites, Quick Reference, Procedure, Pitfalls, Verification) and no nested references. Navigation is easy and the single script reference is clearly signaled. | 5 / 5 |
Total | 19 / 20 Passed |