Content
22%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill is a sprawling, verbose catalog of commands and templates for a specific toolchain (ruv-swarm/claude-flow) rather than a focused, actionable skill. It massively over-explains by listing 8 nearly identical mode descriptions and 8 workflow templates inline, violating token efficiency. Critical workflow safety elements — validation steps, error handling, and feedback loops — are entirely absent despite the skill involving destructive operations like deployments and releases.
Suggestions
Reduce content by 70%+: collapse the 8 modes into a summary table with one representative example, move workflow templates to a separate TEMPLATES.md file, and eliminate duplicated caching/optimization snippets.
Add explicit validation and verification steps to all multi-step workflows, especially deployment and release pipelines (e.g., 'Validate deployment health: gh api /repos/{owner}/{repo}/deployments/{id}/statuses').
Remove descriptive metadata bullets (e.g., 'Coordination Mode: Hierarchical', 'Review Mode: Automated') that provide no actionable value — Claude doesn't need feature marketing copy.
Verify that referenced tools (ruv-swarm, claude-flow@alpha, ruvnet/swarm-action@v1) actually exist and work as described, or clearly mark examples as illustrative templates rather than executable commands.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at 700+ lines. Massive amounts of repetitive content — 8 nearly identical 'mode' descriptions with bullet-point metadata Claude doesn't need, 8 workflow templates that are largely variations of the same pattern, duplicated caching examples, and extensive command listings. Most content describes what tools do rather than providing unique, non-obvious instructions. | 1 / 3 |
Actionability | Commands and YAML snippets appear concrete and copy-paste ready, but they reference tools (ruv-swarm, claude-flow@alpha, ruvnet/swarm-action@v1) that may not exist or be verifiable. The code looks plausible but is essentially pseudocode dressed as real commands — there's no way to confirm these CLIs and actions actually work as shown. | 2 / 3 |
Workflow Clarity | Despite showing multi-step processes (CI/CD pipelines, release workflows), there are no validation checkpoints, error recovery steps, or feedback loops anywhere. The 'self-healing pipeline' template is ironic — it shows no actual validation. Destructive operations like deployment and release have no verification steps. The setup checklist at the end is a checklist, not a validated workflow. | 1 / 3 |
Progressive Disclosure | Uses HTML details/summary tags for collapsible sections, which is a reasonable progressive disclosure technique. However, with no bundle files, all content is monolithically packed into one massive file. The 'Related Skills' section references other skills but they aren't provided. The sheer volume of inline content (8 workflow templates, 8 modes, command reference) should be split into separate files. | 2 / 3 |
Total | 6 / 12 Passed |