check-alignment

Pass

Audited by Gen Agent Trust Hub on Feb 25, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill is restricted to read-only operations on the local codebase and specified documentation paths like /specs. It does not perform any network requests or external data transfers.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface due to scanning untrusted implementation files and documentation. Ingestion points: Scans all .md files in /specs/ and source files in directories like src/, lib/, and app/. Boundary markers: The skill does not define specific delimiters to separate spec content from its own instructions. Capability inventory: Analysis is limited to status reporting and suggesting user-invocable commands; it has no file-write or subprocess execution capabilities. Sanitization: No sanitization is performed on the ingested text. This surface is inherent to its auditing purpose and is considered safe given the lack of dangerous execution sinks.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 25, 2026, 09:41 PM