Content
33%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 extremely lean but provides almost no usable guidance: it restates the description without steps, references, or structure. The real workflow is buried in frontmatter composition rather than surfaced for the reader.
Suggestions
Add a short workflow overview in the body that mirrors the composition steps (brief → clarify → search → research → storyline → pptx → final brief) so the sequence and validation checkpoints are visible.
Reference the orchestrated skills (multi-search-engine, deep-research, pptx) and any expected inputs/outputs at each step so the body is actionable rather than purely descriptive.
Include a brief "what to verify before presenting" checkpoint to satisfy the validation expectation for a multi-step, batch-style generation workflow.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is a single H1 and one sentence with zero padding or restated concepts; it assumes Claude's competence and wastes no tokens. | 5 / 5 |
Actionability | "Builds source-backed presentation packages from a research question" only describes the outcome; it gives no concrete steps, commands, or executable guidance. | 1 / 5 |
Workflow Clarity | The body contains no sequenced steps or validation checkpoints; the actual multi-step workflow lives only in the YAML frontmatter, which is out of scope for this rubric. | 1 / 5 |
Progressive Disclosure | Beyond the H1 there is no structure, no sections, and no signaling to the orchestrated skills (multi-search-engine, deep-research, pptx) or any bundle files; navigation is effectively absent. | 2 / 5 |
Total | 9 / 20 Passed |