Content
57%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 provides concrete, executable code across a clearly sequenced migration workflow with checklists and an error table, but it is verbose, relies on an undefined client, and inlines content that could be split into reference files. Batch operations also lack explicit validation feedback loops.
Suggestions
Add explicit validate-then-proceed checkpoints for the batch lead/account imports (e.g., verify imported counts against source, halt on mismatch) to satisfy the destructive/batch workflow_clarity cap.
Move the long TypeScript implementations into reference files (e.g., references/import-leads.ts) and keep SKILL.md as an overview with one-level-deep links.
Trim the console.log status noise from the code examples and define or link the InstantlyClient so examples are copy-paste ready.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~330-line body avoids explaining concepts Claude already knows, but the TypeScript examples are padded with extensive console.log status prints and could be tightened, fitting the 'mostly efficient but could be tightened' anchor rather than the lean anchor 4. | 3 / 5 |
Actionability | Each step ships concrete, mostly-executable TypeScript (pre-migration audit, account import, CSV lead import, unsubscribe migration, parallel-run monitor), but it depends on an undefined './src/instantly/client' InstantlyClient, leaving minor gaps versus copy-paste-ready anchor 5. | 4 / 5 |
Workflow Clarity | Steps 1-5 are clearly sequenced with a 4-week timeline and cutover checklist, but this batch/destructive migration skill lacks explicit validate-then-proceed feedback loops for the bulk lead/account imports, so the destructive-batch cap holds it at anchor 3. | 3 / 5 |
Progressive Disclosure | No bundle files exist and the document is well-sectioned, but the heavy per-step API implementations are inlined in one ~330-line file rather than split into one-level-deep reference files, fitting the 'some structure, content that should be separate is inline' anchor. | 3 / 5 |
Total | 13 / 20 Passed |