Content
35%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill is well-structured and provides a useful output template, but it's significantly over-explained for Claude's capabilities. Most of the content describes processes Claude already understands (filtering for relevance, analyzing sentiment, pattern recognition). The skill would benefit greatly from being condensed to the output format, source-specific signal thresholds, and the filtering heuristics, cutting the rest.
Suggestions
Cut the content by at least 50% — remove 'When to Use', 'Context Required', and 'Common mistakes' sections which describe things Claude already knows how to reason about.
Add concrete, executable methods for actually accessing these data sources (e.g., specific URLs to fetch, API endpoints, or web search queries to use), since the skill currently only describes what to look for but not how to retrieve it.
Consolidate the workflow into a tighter format: keep the output template and the source-specific signals table, but remove the narrative descriptions of obvious analytical steps like 'filter for relevance' and 'analyze what's working'.
Move the detailed output format template and source signals table to a referenced file, keeping SKILL.md as a concise overview with the core filtering logic and a link to the template.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose for what it conveys. Much of the content describes obvious processes (scan sources, filter for relevance, analyze what's working) that Claude already knows how to do. The 'Context Required', 'Common mistakes', and 'Frameworks' sections largely state the obvious. This could be cut to 1/3 the length without losing actionable value. | 1 / 3 |
Actionability | Provides a clear output template and some concrete guidance (specific platforms, signal thresholds like '500+ upvotes'), but lacks any executable code or API calls for actually fetching data from these sources. The workflow is procedural description rather than concrete commands. The skill essentially describes a thought process rather than providing tools or methods to execute it. | 2 / 3 |
Workflow Clarity | Steps are clearly sequenced (define scope → scan → filter → analyze → generate), but there are no validation checkpoints or feedback loops. No guidance on what to do when sources are inaccessible, when no relevant items are found, or how to verify the accuracy of gathered information. The workflow reads more like a general process description than a precise operational guide. | 2 / 3 |
Progressive Disclosure | References related skills at the bottom (competitive-analysis, market-research, launch-strategy), which is good. However, the main content is a monolithic document that could benefit from splitting — the output template, the source-specific signals table, and the frameworks could be separate references. The inline content is lengthy for a SKILL.md overview. | 2 / 3 |
Total | 7 / 12 Passed |