Content
75%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 dense, highly actionable single-file skill: executable commands, complete CI yaml, and semantically rich exit-code documentation. Its main weaknesses are the missing story-authoring example, no error-recovery guidance tied to exit codes, and all reference-grade detail living inline rather than in bundle files.
Suggestions
Add a minimal Storybook story example showing a parameters.chromatic block (e.g. viewports and delay) so the primary authoring mode is executable too.
Move the full flag table and exit-code catalog to a references/cli.md file, keeping SKILL.md as an overview with well-signaled one-level-deep pointers.
Add brief error-recovery guidance per exit-code family (e.g. on 21-23 Storybook build failures, fix the build locally with --dry-run before re-running).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense with operational detail (flag table, exit-code table, config JSON, CI yaml) and assumes Claude already knows Storybook/Playwright. Minor over-explanation remains, such as the 'built by the team behind Storybook' context line and TurboSnap being explained in three separate places. | 4 / 5 |
Actionability | Install, first-run, flags, config file, and a complete GitHub Actions workflow are copy-paste ready, but the primary authoring path ('Authoring stories (Storybook mode)') describes parameters.chromatic knobs (viewports, delay, pauseAnimationAtEnd) without any story code example. Not 5: a common case lacks executable code. | 4 / 5 |
Workflow Clarity | Install → baseline-establishing first run → author → run → exit-code classification → CI gating is clearly sequenced, and the exit-code table plus the 'fail on exit != 0 (or != 0,1)' guidance act as checkpoints. No destructive/batch cap applies, but there is no explicit error-recovery loop (e.g. what to do on exit 21-23), keeping it below 5. | 4 / 5 |
Progressive Disclosure | Sections are well organized and external doc references are clearly signaled, but the skill is a single 227-line file: the full CLI flag table, exit-code list, and TurboSnap detail are all inlined in SKILL.md rather than split into one-level-deep reference files. Not 5: no reference-file structure; not 3: navigation and signaling are good. | 4 / 5 |
Total | 16 / 20 Passed |