skills/patrickruddiman/skills/slice/Gen Agent Trust Hub

slice

Pass

Audited by Gen Agent Trust Hub on May 3, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill writes generated design documentation to the local file system in a directory derived from the path of a provided specification file (e.g., 'docs/plans/.../slices/').- [DATA_EXFILTRATION]: Phase 3 (Codebase reconnaissance) performs wide-reaching searches using glob and grep tools and reads up to 7 source files to identify existing implementation patterns. This accesses codebase metadata and content to ground implementation choices.- [PROMPT_INJECTION]: The skill processes content from a parent specification file (e.g., 'prd.md'), which creates a surface for indirect prompt injection if that file contains malicious instructions.
  • Ingestion points: Parent specification file read in Phase 1.
  • Boundary markers: Absent; the agent reads the full spec to ground subsequent decisions.
  • Capability inventory: File reads (reconnaissance), repository search (glob/grep), and file writes (markdown documentation).
  • Sanitization: Absent; content from the spec is used directly to drive the design workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
May 3, 2026, 06:25 PM