Content
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A highly actionable skill body with executable examples and a clear scaffold workflow, held back only by mild verbosity and a single-file structure that could offload tangential detail to references.
Suggestions
Move the 'Anthropic / Claude plugin compatibility' section into a reference file and link to it, keeping the core scaffold flow in SKILL.md.
Trim the second full code example to focus on the new concept (system.transform hook + multiple tools) rather than re-demonstrating the zod args pattern already shown in the minimal shape.
Add a verification checkpoint to the scaffold workflow (e.g., 'restart the session and confirm the tool appears') to close the loop on plugin loading.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly lean with no basic-concept padding, but the two overlapping full code examples (minimal shape and multi-tool) and the tangential Anthropic-compatibility table add length that could be tightened. | 2 / 3 |
Actionability | Provides copy-paste-ready executable TypeScript/zod examples, concrete file locations, and explicit opencode.json snippets — fully actionable rather than descriptive. | 3 / 3 |
Workflow Clarity | The 'Scaffold workflow' gives a clear three-step sequence (create file, register in opencode.json, restart session) and this scaffolding task is not a destructive/batch operation requiring feedback loops. | 3 / 3 |
Progressive Disclosure | Single well-sectioned file with clear headings, but at ~160 lines with no external references the compatibility section and long multi-tool example are content that could be split into reference files. | 2 / 3 |
Total | 10 / 12 Passed |