rules-eval
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides a framework for auditing project rules. It defines clear quality standards and provides modules for validating YAML frontmatter and glob patterns. All mentioned scripts and resources, such as
scripts/rules_validator.py, are local vendor resources related to the skill's primary purpose. - [INDIRECT_PROMPT_INJECTION]: The skill analyzes user-created markdown files, which presents a surface for indirect instructions.
- Ingestion points: Markdown rule files within the
.claude/rules/directory, as described inSKILL.md. - Boundary markers: Evaluation logic in
modules/content-quality-metrics.mdandmodules/frontmatter-validation.mdensures that rule content is analyzed against specific metrics rather than executed. - Capability inventory: The skill uses local file system access to read and validate rules and references a local validation script
scripts/rules_validator.py(found inSKILL.md). - Sanitization: The skill implements checks for YAML syntax and glob pattern validity to ensure rule configurations are well-formed.
Audit Metadata