Content
88%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 highly actionable, well-structured CLI skill body with executable commands and a clear validated workflow. The main weakness is minor redundancy and a long inline loop that could be extracted to a reference.
Suggestions
De-duplicate the GraphQL query and jq filter: define each once and reuse it inside the pagination loop rather than restating the full block.
Consider moving the multi-line GraphQL pagination loop into a script under scripts/ (e.g. scripts/unresolved_threads.sh) and referencing it from SKILL.md to shorten the body.
Add a one-line validation checkpoint in the workflow for the destructive resolve-thread action, e.g. re-confirm `isOutdated`/fix-agreement before invoking the resolve mutation.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and free of concept padding, but the GraphQL query and jq filter are each repeated (standalone and inside the pagination loop), which could be tightened. | 4 / 5 |
Actionability | Provides fully executable, copy-paste-ready commands covering the common cases: `gh pr view` status, GraphQL unresolved-thread pagination, jq filtering, and the resolve-thread mutation. | 5 / 5 |
Workflow Clarity | A clearly sequenced 7-step workflow with an explicit feedback loop (step 5: fix, run tests, push, return to step 2) and concrete stop criteria in step 7, including verification before resolving stale threads. | 5 / 5 |
Progressive Disclosure | Well-organized into Workflow, GitHub CLI Checks, and Operating Rules sections with no nested references, but the long inline pagination loop is content that could live in a separate reference file. | 4 / 5 |
Total | 18 / 20 Passed |