cut
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEDYNAMIC_CONTEXT_INJECTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DYNAMIC_CONTEXT_INJECTION]: The skill utilizes the dynamic context injection pattern to execute a local configuration script when initialized.\n
- Evidence: The frontmatter contains a
!bashcommand that callsyaml-helper.shto resolve project-specific automation settings.\n- [INDIRECT_PROMPT_INJECTION]: The skill reads from project files that could potentially contain untrusted or malicious instructions, creating a surface for indirect prompt injection.\n - Ingestion points: The skill reads
design/game-brief.md,feel-log.md, and other project metadata.\n - Boundary markers: Human-in-the-loop review is enforced via the
AskUserQuestiontool before any cuts are applied or dates are set.\n - Capability inventory: The skill has the ability to execute file modifications and git commands via
Bashthrough pre-defined local scripts.\n - Sanitization: No explicit sanitization or validation of the ingested markdown content is performed before processing.\n- [COMMAND_EXECUTION]: The skill executes local shell scripts and development tools to manage project state.\n
- Evidence: The skill calls
cut-read.sh,cut-apply.sh,npm, andgodotwith various parameters to read and update project status.
Audit Metadata