Content
78%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.
A lean, information-dense instruction skill: nearly every sentence carries ReMe-specific operational detail Claude could not know, with concrete parameters and premature-conclusion guards throughout. Its main gap is that the core search/read flow is never shown as a concrete example, and no error-recovery loop is spelled out.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~35-line body is dense with non-obvious operational facts (hook batching, config location, pagination semantics) and explains nothing Claude already knows, but a few spots could be trimmed — "Use this skill for focused recall or connection troubleshooting" restates the description and phrasing like "subject to a bounded shutdown budget" is convoluted. | 4 / 5 |
Actionability | Gives concrete parameters throughout — "limit=5", "page=1 and page_size=20; omit date", "Call health_check and version", the copy-pasteable `reme start workspace_dir=/absolute/path/to/workspace service.backend=http` — but defers exact search/read tool names ("discover the actual tool names instead of assuming a fixed namespace") without a single example invocation. | 4 / 5 |
Workflow Clarity | Sequences are present with explicit checkpoints — "Follow total_pages", "If you stop before the last page, say the review is partial", "Check the plugin and /mcp status before concluding the service is stopped", "If a search is empty, say there is no relevant memory" — but there is no explicit validate-and-retry recovery loop and the steps live in prose rather than a visible sequence. | 4 / 5 |
Progressive Disclosure | The skill is under 50 lines, needs no external references, and is organized into two clean sections (Recall, Status and recording) that match its two use cases, which merits a 5 under the rubric's simple-skill provision. | 5 / 5 |
Total | 17 / 20 Passed |