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 body is well-organized with a clear resource map and mode structure, but it leans heavily on reference files and scripts that are not present in the bundle, and its workflows lack explicit validation feedback loops. It reads as a routing overview rather than a self-contained, executable guide.
Suggestions
Ship the referenced bundle (references/HARNESS.md, references/commands/*.md, references/guides/, scripts/repl_skin.py, etc.) so the signaled one-level references actually resolve.
Add explicit validate-then-retry feedback loops to the Build and Test workflows, especially around destructive/batch operations like session mutations and installed-command subprocess coverage.
Replace the directory-tree structure block with concrete, copy-pasteable scaffolding commands or point to a real template file that exists in the bundle.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Sections are organized and avoid explaining basics, but the six-step fallback ladder and the document path-remapping table add length that could be tightened. | 2 / 3 |
Actionability | Concrete specifics exist (find_namespace_packages, CLI_ANYTHING_FORCE_INSTALLED=1, --dry-run, _resolve_cli()), but core build steps delegate to reference files and the structure block is a directory tree rather than executable code. | 2 / 3 |
Workflow Clarity | Modes are clearly delineated with an ordered read-first sequence, but build/test workflows lack explicit validate-then-fix feedback loops; per the rubric, missing feedback loops for batch/destructive ops caps this at 2. | 2 / 3 |
Progressive Disclosure | The body is well structured with a Resource Map and clearly signaled one-level references, but the referenced files (references/*, scripts/*.py, scripts/templates/*) are absent from the bundle and the fallback ladder reaches for an external repo, leaving navigation partly broken. | 2 / 3 |
Total | 8 / 12 Passed |