straymark-audit-execute

Run a read-only security audit by invoking the audit execution from an auditor-side CLI, which consumes a resolved Charter audit-prompt and strictly enforces read-only behavior during the tool-using review process.

Generate a schema-based audit report by executing the audit prompt, requiring every finding to cite exact path:line evidence from actual file reads, and writing the final output to a canonical report path following audit-output.schema.v0.json.

A Charter audit is a standardized security review process that requires evidence-citation in path:line format, ensuring every finding references exact lines from actual file reads to prevent fabricated results.

Yes, you can run parallel audits across multiple model families by auto-selecting pending prompts per auditor model slug, generating model-scoped reports for each, and later consolidating them via a review skill.

Yes, the audit execution supports multi-CLI audit cycles by allowing operators to run multiple model families in parallel, each writing a model-scoped report to a canonical report path for later consolidation.