Content
93%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.
A tight, executable, well-organized skill body that assumes Claude's competence and adds only non-obvious guidance (anti-patterns, staleness caveats). The only slight gap is the absence of explicit validation checkpoints, which is largely inherent to a read-only query skill.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and information-dense throughout — no padding, no explaining basic concepts (AST, functions) Claude already knows; every section (Tools, Workflow, Anti-patterns) earns its tokens with non-obvious, actionable guidance. | 5 / 5 |
Actionability | Fully executable tool invocations with realistic example parameters (hivemind_graph_search({ pattern: 'auth+handler' }), neighborhood for 'src/hooks/capture.ts') and a concrete config path, covering the common search-by-symbol and neighborhood-by-file cases. | 5 / 5 |
Workflow Clarity | Clear 3-step decision sequence (search → neighborhood → read source_file:line) with soft retry hints in Anti-patterns ('try multiple keywords if the first misses', 'prefer live source'), but no explicit validation checkpoints since it is a read-only query skill. | 4 / 5 |
Progressive Disclosure | Self-contained ~55-line skill with no need for external references, organized into well-signaled sections (When to use, Tools, Workflow, When NOT to use, Anti-patterns); nothing that belongs in a separate file is inlined. | 5 / 5 |
Total | 19 / 20 Passed |