Content
86%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 well-structured, lean workflow document: it front-loads a deterministic script call, sequences browser automation steps clearly, and correctly references its bundled script. The main weaknesses are descriptive rather than executable UI-locator guidance and the absence of error-recovery branches for the fragile browser-automation steps.
Suggestions
Add fallback guidance for fragile UI steps — e.g., what to do if the 'Crop' selector or caption area is not found, such as likely alternative locations or aria-labels to try — to move actionability from descriptive to executable.
Add a short error-recovery loop after publishing (e.g., if the share confirmation does not appear within a reasonable wait, re-check for the post on the profile or retry the Share click) to strengthen workflow_clarity.
Optionally include an example selector or two (e.g., `input[type='file']`, the Create button's aria-label) so agents can copy-paste locator strategies instead of interpreting prose descriptions.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~35-line body is lean and every line is instructional — it opens with a two-sentence overview, a concrete script invocation, and numbered browser steps — with no padding or explanations of concepts Claude already knows. This matches the 'lean and efficient; every token earns its place' anchor. | 5 / 5 |
Actionability | There is a fully executable command (`python scripts/get_media_aspect_ratio.py <path_to_media_file>`) with enumerated outputs, plus concrete technique guidance like setting the hidden `<input type="file">` element. It falls short of 5 because the UI steps are descriptive rather than executable — "usually located at the bottom left", "Locate and click" — with no example selectors, and no handling for variations in Instagram's UI. | 4 / 5 |
Workflow Clarity | The four steps are clearly sequenced with data flowing across them (Step 1's crop ratio feeds Step 2's selection) and an end checkpoint ("Await confirmation that the post has been shared successfully"). It is not a 5 because there are no error-recovery feedback loops — no guidance for what to do when the upload fails, the crop button is missing, or confirmation never appears. | 4 / 5 |
Progressive Disclosure | The skill is under 50 lines with no need for external references, and the one bundle file it does reference (`scripts/get_media_aspect_ratio.py`) exists on disk, is referenced one level deep with the exact invocation and expected outputs. Sections are well organized, which per the simple-skill guidance merits a 5. | 5 / 5 |
Total | 18 / 20 Passed |