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 a dense, well-organized routing skill with concrete directives, a clear execution order, and an accountability checkpoint for skipped steps, practicing the token efficiency it mandates. Its main weakness is progressive disclosure against the real bundle: all 16 referenced playbooks are missing and the one shipped reference file is orphaned.
Suggestions
Ship the referenced playbook files (or correct the paths) so the 16 `playbooks/*.md` references resolve within the bundle.
Link `references/plan.md` from the body (e.g., from the Playbooks section as the plan playbook) or remove it, since it is currently unreachable from SKILL.md.
Move the long AskUserQuestion-classification bullet's rationale into the prototype/investigation playbooks to tighten the Non-negotiables section.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and imperative with no explanations of concepts Claude already knows ("**Short declarative sentences.** One thought per sentence, ended with a period.") and practices the brevity it preaches; a few multi-clause passages could be trimmed ("The ask is the slow path. A throwaway probe usually answers faster, and it hands the human a result to react to instead of a decision to make."). | 4 / 5 |
Actionability | Concrete, executable directives throughout: exact strings ("Use `subagent_type: "pstack:poteto-agent"`"), model defaults ("defaults `sonnet` for code, `opus` for prose and judgment"), a worked example ("Write `assert(ok, 'persisted across restart')`"), and the PR URL format; the gap is that much of the executable detail delegates to playbook and leaf-skill files not shipped in this bundle. | 4 / 5 |
Workflow Clarity | Sequence is explicit ("Your first todolist actions are the matched playbook's steps, copied in verbatim, before any task-specific todos and before you reason about the task") with checkpoint enforcement ("A step you choose not to do stays in the list with a one-line `skip: <reason>`; skipping silently is not allowed"); not 5 because the execution-level step sequences and validation loops live in the playbook files rather than the body. | 4 / 5 |
Progressive Disclosure | References are clearly signaled and one level deep ("`playbooks/feature.md`", "`playbooks/investigation.md`"), but scored against the actual bundle: none of the 16 referenced `playbooks/*.md` paths exist in the bundle, and the single shipped file, `references/plan.md`, is never linked from the body, so navigation does not fully resolve. | 3 / 5 |
Total | 15 / 20 Passed |