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, highly actionable skill with clear workflow routing between pre-commit and default scan paths, executable commands, and explicit error handling. Its main weakness is moderate verbosity—the secret types reference table and duplicated rotation guidance across sections add tokens without proportional value. The presentation templates are thorough but lengthy for inline inclusion.
Suggestions
Consider moving the 'Secret Types Detected' table to a separate reference file since Claude doesn't need it to execute the scan workflow—it's informational context that inflates the token cost.
Consolidate the rotation guidance which appears in both 'Immediate Actions' and 'Recommendations' into a single concise block to reduce redundancy.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is mostly efficient and domain-specific, but includes some redundancy—the 'Secret Types Detected' table is informational padding Claude doesn't need to act on, and the shared rules/recommendations sections repeat rotation guidance multiple times with slightly different phrasing. Some tightening is possible. | 2 / 3 |
Actionability | Provides fully executable commands (npx -y endorctl scan with specific flags), concrete MCP tool parameters (scan_types, scan_options), specific resource_type values for hydration, and copy-paste-ready markdown templates for output presentation. The routing logic is explicit and the CLI fallback is clearly specified. | 3 / 3 |
Workflow Clarity | Two clearly separated workflows (pre-commit vs default) with explicit routing criteria, numbered sequential steps, validation/hydration steps, and a comprehensive error handling table covering both paths. The 'do not' guardrails (e.g., don't use MCP scan for pre-commit, don't use --output-type with --pre-commit-checks) serve as validation checkpoints preventing common mistakes. | 3 / 3 |
Progressive Disclosure | References external files (rules/endor-safety.md, references/data-sources.md, CLAUDE.md, /endor-setup) which is good, but the skill itself is quite long (~150+ lines) with the full presentation templates inlined. The two presentation templates and the secret types table could potentially be split into referenced files. However, no bundle files are provided, so we can't verify the references resolve. | 2 / 3 |
Total | 10 / 12 Passed |