Content
50%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The skill is highly actionable with concrete, executable commands throughout, but it is significantly over-verbose by essentially reproducing the gh CLI's documentation that Claude already knows. The progressive disclosure structure exists but is underutilized—most reference content is inline rather than in separate files. Workflow examples include some verification steps but lack systematic error recovery patterns.
Suggestions
Remove or drastically reduce the 'Commands (Direct gh Usage)' sections—Claude already knows gh CLI commands. Focus only on the wrapper script usage and the read/write distinction, which is the novel information.
Move the Advanced Usage, Configuration, and Troubleshooting sections to reference files and link to them from the main SKILL.md to improve progressive disclosure.
Add explicit error handling/recovery steps to workflows involving destructive operations (merge, close, delete), e.g., 'If merge fails due to conflicts: ...'
Remove the Summary section which repeats the Prerequisites and Script Usage sections almost verbatim.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is very verbose, essentially duplicating the gh CLI's own help documentation. Claude already knows how to use `gh` commands like `gh issue list`, `gh pr create`, etc. The sections on direct gh usage, configuration, aliases, JSON output, and API access are all things Claude knows well. The summary section repeats what was already stated. Much of this content adds no new knowledge beyond the wrapper script usage. | 1 / 3 |
Actionability | The content provides fully executable, copy-paste ready commands throughout. Every section includes concrete bash commands with specific flags and arguments. The examples section shows complete multi-step workflows with real command sequences. | 3 / 3 |
Workflow Clarity | The examples section shows multi-step workflows with some verification steps (e.g., 'Verify approval was recorded', 'Verify PR was created and linked'), which is good. However, there are no explicit error recovery loops or validation checkpoints for destructive operations like closing issues, merging PRs, or deleting repos. The distinction between read (wrapper) and write (direct gh) operations could be clearer in the workflow context. | 2 / 3 |
Progressive Disclosure | There are references to `references/permissions.md` and `references/common-workflows.md`, which is good progressive disclosure. However, the main SKILL.md contains a large amount of reference-style content (full command listings for issues, PRs, workflows, repos, search, configuration, troubleshooting, API access) that could be offloaded to reference files. The inline content is too heavy for an overview document. | 2 / 3 |
Total | 8 / 12 Passed |