Content
71%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 highly actionable and exceptionally well-sequenced with strong validation checkpoints, but it leans verbose and keeps most procedural detail inline rather than progressively disclosing it to reference files.
Suggestions
Tighten the long prose paragraphs (validation surface, guarded publication steps) into terser checklists to reduce tokens without losing the guard semantics.
Move the detailed publication-guard and validation-surface procedures into a dedicated reference file, keeping SKILL.md as an overview with one-level-deep links.
Consolidate the repeated stop/condition language — many near-identical retry-boundary statements appear across sections and could be stated once.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is densely verbose — long procedural paragraphs and heavily repeated stop/guard language (e.g. the multi-line validation-surface enumeration and publication-guard steps) add padding beyond what an intelligent reader needs, though it stays accurate. | 2 / 5 |
Actionability | Provides concrete, copy-paste-ready shell commands (gh API calls, git fetch/rev-parse, npm run validate:pr) plus precise conditional-update guard steps that cover the common cases. | 5 / 5 |
Workflow Clarity | Multi-step publication process is clearly sequenced with explicit validation checkpoints and feedback loops (re-fetch base, re-read headRefOid, retry/recovery boundaries, stop conditions) appropriate to a destructive batch operation. | 5 / 5 |
Progressive Disclosure | Content is organized into sections and points to shared references (../_shared/pr-follow-up.md, git-github-hard-stop.md, documentation-writing-review.md), but the bulk of the procedural detail is inlined in SKILL.md rather than split into separate files. | 3 / 5 |
Total | 15 / 20 Passed |