rjv-spec-driven
Pass
Audited by Gen Agent Trust Hub on Aug 19, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a Python script,
scripts/check_spec_coverage.py, which is designed to be executed by the agent or user to validate that acceptance criteria IDs defined in specification files are referenced within test suites. The script performs localized file reading and regex-based searching with proper escaping of inputs. - [DATA_EXPOSURE]: The skill requires the agent to interact with project files, including creating and updating documentation in directories like
_docs/features/anddocs/adr/. This access is standard for development tools and is restricted to the local project context as defined by the skill's operational purpose. - [PROMPT_INJECTION]: The instructions include directives such as a "one-shot rule" to follow specific templates exactly. These are architectural constraints for output formatting rather than attempts to bypass safety filters or override system-level instructions.
Audit Metadata