Content
92%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 is a lean, well-structured, instruction-only skill with a clear sequenced workflow, explicit verification checkpoints, and valuable non-obvious pitfalls. Its only gap is the absence of concrete example commands, which keeps actionability at 4 rather than 5.
Suggestions
Add one or two concrete example read-only probes (e.g., "which <tool>", "ls -l <path>", checking exit code with "echo $?") to close the minor actionability gap.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~27-line body is entirely directive — "Read the command, exit status, standard output, and standard error", "Do not rerun an unchanged failing command repeatedly" — with zero padding and no explanation of concepts Claude already knows. Every token earns its place, matching anchor 5. | 5 / 5 |
Actionability | Guidance is concrete and executable: "Classify the failure as syntax, missing dependency, wrong path, permissions, resource pressure, timeout, service state, or task logic" and "use the returned session identifier to poll or provide input" give specific, followable directives. It falls short of anchor 5 only because no concrete example commands or common-case illustrations are included (e.g., example read-only probes), leaving minor gaps — anchor 4's profile. | 4 / 5 |
Workflow Clarity | A clearly sequenced 6-step procedure (gather evidence → probe environment → classify → change one condition and retry → manage long-running sessions → verify) with a dedicated "## Verification" section defining explicit acceptance checks ("The corrected command exits as expected", "passes an independent acceptance check"). Step 4 embodies an explicit feedback loop (diagnose → change one condition → retest), and the Pitfalls section guards against blind reruns — matching anchor 5. This is not a destructive/batch operation, so no cap applies. | 5 / 5 |
Progressive Disclosure | The body is under 50 lines with four well-organized sections (When to Use, Procedure, Pitfalls, Verification), no bundle files, and no content that needs splitting into references. Per the rubric's simple-skill guideline, this earns anchor 5 on well-organized sections alone. | 5 / 5 |
Total | 19 / 20 Passed |