Content
100%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.
This is a model single-purpose skill: lean and tool-specific prose, a copy-paste executable guard-and-command block with expected output, explicit failure and verification handling, and clean section organization. There is no padding, no nested references, and no missing validation for the operation it performs.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is ~44 lines with no padding: the opening paragraph conveys non-obvious, tool-specific behavior (lane inference precedence and why a finished dirty-tree session still reads as working), and every Note is an actionable specific (auto-release semantics, one-attempt policy, accepted status values, verify command). Nothing explains concepts Claude already knows, matching the 'every token earns its place' anchor. | 5 / 5 |
Actionability | Fully executable, copy-paste ready: `[ -n "$CMUX_WORKSPACE_ID" ] || exit 0` guard followed by `cmux workspace status set done`, with expected output, explicit failure handling ("If the command exits non-zero ... ignore it. Do not retry"), and a verify command (`cmux workspace status`). Covers the common cases including the wrong-workspace failure mode. | 5 / 5 |
Workflow Clarity | Single-task skill where the action is unambiguous, which the guidelines say can score 5: check the guard, run the command, compare against expected output, and confirm with `cmux workspace status`. This is not a destructive or batch operation, so no validation cap applies; failure behavior is explicitly specified. | 5 / 5 |
Progressive Disclosure | Under 50 lines, single task, no bundle files exist or are needed; content is organized into well-labeled sections (When to invoke, Guard and command, Notes) with the overview up front. Per the guidelines, a skill of this size with no need for external references scores 5 on well-organized sections alone. | 5 / 5 |
Total | 20 / 20 Passed |