Content
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured troubleshooting skill with clear workflow steps, concrete resolution patterns, and good actionability. The ecosystem detection table and structured diagnosis output template make it immediately usable. Minor weaknesses include some inline content that could be offloaded to reference files and the inability to verify referenced bundle files exist.
Suggestions
Verify that references/error-knowledge-base.md and references/data-sources.md exist in the bundle, or include them — the skill's core matching logic (Step 3) depends entirely on the knowledge base file.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Generally efficient with good use of tables and structured formatting, but some sections like the ecosystem detection table and common resolution patterns could be tighter. The output template includes template variables and conditional blocks that add bulk but serve a purpose. | 2 / 3 |
Actionability | Provides concrete, specific guidance throughout: exact MCP tool invocations with parameters, specific file paths to check (~/.endorctl/config.yaml), specific env var names, exact commands (rm -rf ~/.endorctl), links to relevant docs, and clear resolution patterns per category. The diagnosis output template is copy-paste ready. | 3 / 3 |
Workflow Clarity | Clear 5-step sequential workflow (detect ecosystem → classify → match patterns → present diagnosis → handle multiple errors) with explicit priority ordering for multiple errors, a feedback loop via re-scan after fix, and an error handling table covering edge cases. The auth conflict resolution provides a clear decision tree. | 3 / 3 |
Progressive Disclosure | References external files (references/error-knowledge-base.md, references/data-sources.md) and other skills (/endor-setup, /endor-scan) appropriately, but no bundle files were provided to verify these exist. The main SKILL.md contains a good amount of inline content that could potentially be split out (e.g., the full ecosystem detection table, common resolution patterns), though it's not egregiously long. | 2 / 3 |
Total | 10 / 12 Passed |