feature-design

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses standard shell commands to facilitate a development workflow, specifically invoking date for document timestamps and git commit/git push to synchronize specification documents with a remote repository.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from existing project documentation and user-provided feature ideas to generate new specifications.
  • Ingestion points: Reads project context from .monkeycode/docs/INDEX.md, .monkeycode/docs/ARCHITECTURE.md, and existing files in .monkeycode/specs/.
  • Boundary markers: None explicitly defined for isolating content during file ingestion.
  • Capability inventory: Performs filesystem writes to create new requirement and design files, and executes Git commands for persistence.
  • Sanitization: No explicit sanitization or validation of the input documentation is specified beyond formatting it into EARS patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 03:36 PM
Security Audit — agent-trust-hub — feature-design