Content
50%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 hive-mind role prompt with a clear section structure and a consistent (if templated) reporting protocol, but it is built from non-executable placeholder pseudocode repeated across seven near-identical blocks, includes a malformed duplicate frontmatter, and offers no progressive disclosure or validation loops. It functions as inspiration for a persona rather than as operational skill guidance.
Suggestions
Collapse the seven repetitive memory_usage blocks into one parameterized example (with the full key/value schema documented once) and vary only the parts that differ per pattern.
Make the code executable or explicitly justify the template form: replace placeholders like "scout-[ID]" and "what was found" with concrete fill-in instructions or valid syntax.
Move the per-pattern exploration templates (codebase/dependency/performance scout) into a references/ file and add a validation step (e.g., confirm the store succeeded before proceeding) to the reporting workflow.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Sections like "Scouting Strategies", "Integration Points", and the Do/Don't lists are tight, but seven near-identical mcp__claude-flow__memory_usage template blocks padded with dummy data ("total_count: 45", "load_time_ms: 3500") repeat the same call structure, and the stray duplicate frontmatter block at the top of the body is pure noise. This sits between the 'mostly efficient' (3) and 'noticeably verbose' (2) anchors — the padding is substantial but there is little conceptual over-explanation. | 3 / 5 |
Actionability | Concrete guidance exists — real tool name, key names, namespaces, and value schemas are specified — but the blocks are non-executable templates: invalid JS object literals, unfilled placeholders like "scout-[ID]", "what was found", and "reconnaissance type". This matches the 'pseudocode instead of executable code; missing key details' anchor, not 4 which requires mostly executable guidance. | 3 / 5 |
Workflow Clarity | The three Scouting Strategies give numbered, legible sequences and "Verify findings before alerting" appears in the Do list, but the deploy → discover → report protocol has no real validation checkpoints, error-recovery, or feedback loops around the memory operations — matching the 'steps listed but checkpoints missing or implicit' anchor. | 3 / 5 |
Progressive Disclosure | The body has consistent ##/### section headers and is navigable, but it is a 247-line monolith with no reference files anywhere in the bundle (references/, scripts/, assets/ absent) and no in-body pointers, while the repetitive per-pattern call templates are exactly the content that belongs in a separate reference file — matching 'some structure but could be better organized'. | 3 / 5 |
Total | 12 / 20 Passed |