Content
85%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.
Highly actionable and well-sequenced content with strong validation checkpoints and exemplary progressive disclosure verified against the actual bundle. The main weakness is redundancy: several safety and no-op rules are restated across multiple sections, inflating token cost without adding guidance.
Suggestions
State the mission-change confirmation and workspace-reset rules once (e.g., in Constraints And Safety) and reference them from the dispatch table and Gotchas instead of repeating them in four sections.
Merge the no-op/blocker rule from When To Use with the 'incomplete until a durable artifact or blocker' rule in Artifact-specific requirements into a single statement.
Tighten the Philosophy section's abstract phrasing (e.g., 'knowledge from trusted sources, skills through practice loops, and wisdom through real-world feedback') into the concrete operating rules it already restates in the Procedure.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient and dense, but guidance is repeated across sections: mission-change confirmation appears in the Procedure, the dispatch table, Constraints And Safety, and Gotchas; the no-op/blocker rule is stated in both When To Use ("do not return an empty or no-op result") and Artifact-specific requirements ("incomplete until Teach writes a durable workspace file or durable blocker note"). It is not 2 because there is no padding or explanation of concepts Claude already knows; not 4 because the repetition goes beyond 'minor instances that could be trimmed'. | 3 / 5 |
Actionability | The body provides a copy-paste-ready blocker markdown example, exact artifact paths in the dispatch table, executable validation commands (./bin/ask skills audit/evals/external-review), and per-request evidence requirements — fully actionable instruction guidance with common cases covered in the Examples section. As an instruction-only skill, absence of code is not penalized per the scoring notes. | 5 / 5 |
Workflow Clarity | The 9-step Procedure is clearly sequenced with explicit validation gates ("Verify created paths and links before closeout", fail-fast Validation section), feedback loops for error recovery (blockers, "fix the smallest failing scope before continuing"), and confirmation gates for destructive mission overwrites and workspace resets. | 5 / 5 |
Progressive Disclosure | The body is a clear overview with well-signaled, one-level-deep references — every linked file (contract.yaml, evals.yaml, task-profile.json, templates.md, teaching-operating-model.md, the four format files, knowledge-capsule-routing.md, knowledge-capsule.manifest.yaml) exists in references/. A dedicated Progressive Disclosure section directs when to read each file, and capsule routing prevents loading all capsules by default. | 5 / 5 |
Total | 18 / 20 Passed |