Content
56%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 CLI usage guide is genuinely strong — executable commands, a well-sequenced connection workflow with state polling and error recovery, and useful best practices. The skill is dragged down by a massive, mostly-irrelevant bullet 'overview' that inflates token cost and belongs in a separate file or should be removed entirely.
Suggestions
Delete or externalize the 'Overledger Overview' bullet list: the vast majority of entries (Spiritual Growth, Venture Capital, Climate Change, etc.) are unrelated to Overledger and add no actionable value.
Fix the broken step references in the connection flow — label the 'Searching for actions' step as 'Step 2' and reconcile the stray '1b' heading so the sequence is unambiguous.
De-duplicate action discovery: the 'Searching for actions' and 'Popular actions' sections repeat the same `action list --intent` command; consolidate them and add a brief note on interpreting inputSchema/outputSchema.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The actionable sections are reasonably lean, but the ~290-line 'Overledger Overview' bullet list (e.g. 'Spiritual Growth', 'Venture Capital', 'Initial Public Offering', 'Climate Change') is heavy, largely irrelevant padding that does not earn its tokens, fitting the 'noticeably verbose; several padded sections' anchor and approaching the severely-padded level 1. | 2 / 5 |
Actionability | It provides concrete, executable commands throughout (npm install, membrane login, connection ensure/get --wait, action list/run, request with a full flag table), covering the common cases copy-paste ready; minor gaps are unresolved placeholders (CONNECTION_ID, <actionId>) and a redundant 'Popular actions' section repeating the search command. | 4 / 5 |
Workflow Clarity | The connection flow is clearly sequenced with state-based validation checkpoints and a feedback loop (poll --wait, handle CLIENT_ACTION_REQUIRED, re-poll after user action, check CONFIGURATION_ERROR); it stays at 4 rather than 5 because 'Step 2' is referenced but unlabeled, the '1b' numbering is inconsistent, and action-run results lack a verify step. | 4 / 5 |
Progressive Disclosure | The body has clear section headers, but with no bundle files and everything inlined, the ~290-line overview list is content that should live in a separate reference file or be deleted, matching the 'some structure but content that should be separate is inline' anchor; the under-50-line exception does not apply to this ~440-line skill. | 3 / 5 |
Total | 13 / 20 Passed |